Skip to main content

Full text of "The Journal of Biological Chemistry"

See other formats


Google 



This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 



OF TFF 
UN/Vi kxjVy 



Chemistry Lib* 



i 



. ' 



•I 



\ 



if 



. THE JOURNAL 




n 4 



OF 



BIOLOGICAL CHEMISTRY 

roUKDED DT CBRIHTIAK A. HBRTSR AKD SUSTAINED IN PART BY THE CHRISTIAN A. HERTER 

MEMORIAL FUND 



EDITED B7 



a D. DAKIN, New York City. LAFAYETTE B. MENDEL, New Haven, Conn. 

E. K. DUNHAM, New York City. A. N. RICHARDS, Philadelphia, Pa. 

DONALD D. VAN SLYKE, New York City. 



WITH THE COLLABORATION OP 

J. J. ABEL, Baltimore, Md. J. B. LEATHBS, Toronto, Cansda. 

R. H. CHITTENDEN. New Haren. Conn. P. A. LEVENB, New York. 

OTTO FOLIN. Boston, Mass. JACQUES LOEB. New York. 

WILLIAM J. GIES, New York. A. S. LOEVENHART, Madison. Wis. 

L. J. HENDERSON, Cambridge, Mass. GRAHAM LUSK, New York. 

REID HUNT, Boston, Mass. A. B. MACALLUM, Toronto, Canada. 

W. A. JACOBS, New York. J. J. R. MACLEOD, CleveUnd, Ohio. 

WALTER JONES, Baltimore. Md. JOHN A. MANDEL, New York. 

J. H. KASTLE. Lexington, Ky. A. P. MATHEWS, Chicago, lU. 

F. G. NOVY, Ann Arbor, Mich. 

THOMAS B. OSbOKNK, ^er' R-'.fen, Conn^ 

T. BRAILSFORD !l6B£Kl^|^|0AleQp«. 

P. A. SHAFl - -* ^ 

A. E. TAYLOl 

F. P. UNDERH^, NiMHIVeii,Conn. 

V. C. VAUGHAN, Ann Arbor. Mich. 

ALFRED J. WAKEMAN. New Haven, Conn. 




oiln'KriajjiphMy 

CRH^i^.M^Hive] 



\^K ^ 



VOLUME XX 
Baltimore 



• / •• • • . ; . 









COPYRIGBT igi5 

■T 

THE JOURNAL OF BIOLOGICAL CHEMISTRY 



Chomlatry Lii>« 



% s 



I • ' 






. f 



• * 



PUBLISHED BT THE ROCKEFBLLBR INSTITUTE FOB MEDICAL BE8BABCH FOB THE 

JOURNAL OF BIOLOGICAL CHBIHISIBT, INC. 

COMPOSED AND PRINTED AT THE 

WAVERLY PRESS 
Bt the Wiluams db Wiletns Company 
Baltimore, U. S. A. 



"sm^ 



0)P€O\ 




CHEMISTRY 
LIBRARY 



CONTENTS OF VOLUME XX. 

W. B. Thompson: Studies in the blood relationship of animals as 
displayed in the composition of the serum proteins. III. A 
comparison of the sera of the hen, turkey, duck, and goose with 
respect to their content of various proteins 1 

R. S. Briggs: Studies in the blood relationship of animals as dis- 
played in the composition of the serum proteins. IV. A com- 
parison of the sera of the pigeon, rooster, and guinea fowl with 
respect to their content of various proteins in the normal and in 
the fasting condition 7 

Paul J. Hanzlik: Precipitation of serum-albumin and glutin by al- 

kaloidal reagents 13 

Walter Jones and A. E. Richards : Simpler nucleotides from yeast 

nucleic acid ' 25 

Sergius Morgulis : Studies on fasting flounders 37 

Joseph C. Bock and Stanley R. Benedict: An estimation of the 

Folin-Farmer method for the«colorimetric estimation of nitrogen. 47 

Robert C. Lewis and Stanley R. Benedict: A method for the esti- 
mation of sugar in small quantities of blood 61 

F. B. Kingsbury and E. T. Bell: The synthesis of hippuric acid in 

experimental tartrate nephritis in the rabbit 73 

Arthur W. Dox: The soluble polj'saccharides of lower fungi. III. 
The influence of autolysis on the mycodextran contenf of As^t" 

gillus nigcr 83 

C. G. Imrie: On the fat in the blood in a case of lipaemia 87 

A. W. Bosworth : Fibrin 91 

P. A. Levene and F. B. La Forge: Correction. On chondroitin 

sulphuric acid 95 

Robert M. Chapin and Wilmer C. Powick: An improved method 
for the estimation of inorganic phosphoric acid in certain tis- 
sues and food products 97 

Frederic Fenger: On the size and composition of the thymus gland. 115 
Andrew Hunter and Sutherland Simpson: The influence of a diet 

of marine algae upon the iodine content of sheep's thyroid 119 

Jacob Rosenbloom: A note on the distribution of mercury in the 

body in a case of acute bichloride of mercury poisoning 123 

G. W. Raiziss and H. Dubin: On the estimation of benzoic acid in 

urine 125 

R. T. Woodyatt: Studies on th^ theory of diabetes. IV. The paral- 
lelism between the effects of the pancreas and those of metallic 
hydroxides on sugars 129 

• • • 

111 



M580W± 



IV ., Copteats .- 

Lucius L. Van Slyke and Alfred W. Boswokth: Condition of 

casein and salts in milk 135 

Carl O. Johns and Byron M. Hendrix: Researches on purines. 
XVI. On the isomeric monomethyl derivatives of 2-methyl- 
mercapto-4-amino-6-oxypyrimidine. On l-methyl-2-methylmer- 
capto-6,8-dioxy purine 153 

C. G. MacArthur and C. L. Luckett: Lipins in nutrition 161 

Reuben L. Hill: Note on the use of colloidal iron in the determina- 
tion of lactx)se in milk 176 

David Fraser Harris and Henry Jermain Maude Creighton: 

Spectroscopic investigation of the reduction of hemoglobin by 
tissue reductase 179 

Frank P. Underhill and Albert G. Hogan: Studies in carbohy- 
drate metabolism. VIII. the influence of hydrazine on the 
utilization of dextrose 203 

Frank P. Underhill and Albert G. Hogan: Studies in carbo- 
hydrate metabolism. IX. The influence of hydrazine on the 
glyoxalase activity of the liver 211 

Victor John Harding and Reginald M. MacLean: A colorimet- 

ric method for the estimation of amino-acid a-nitrogen 217 

Francis G. Benedict and Paul Roth: The metabolism of vege- 
tarians as compared with the {netabolism of non-vegetarians 
of like weight and height 231 

Francis G. Benedict and H. Monmouth Smith: The metabolism 
of athletes as compared with normal individuals of similar 
height and weight 243 

F. G. Benedict and L. E. Emmes: A comparison of the basal metab- 
olism of normal men and women 253 

Francis G. Benedict: Factors affecting basal metabolism 263 

Francis G. Benedict: A respiration apparatus for small animals 3Q1 

\V. T. Bovie: Simple quartz mercury-vapor lamps for biological 

and photochemical investigations 315 

/^'-N. W. Janney: The metabolic relationship of the proteins to glucose. 321 
/ Thomas B. Osborne and Lafayette B. Mendel (with the cooperation 
of Edna L. Ferry and Alfred J. W'akeman): The compara- 
tive nutritive value of certain proteins in growth, and the prob- 
lem of the protein minimum 351 

Thomas B. Osborne and Lafayette B. Mendel: Further obser- 
vations of the influence of natural fats upon growth 379 

Victor C. Myers and Morris S. Fine : The non-protein nitrogenous 
compounds of the blood in nephritis, with special reference to 
creatinine and uric acid 391 

Mary Louisk Foster: Studies on a method for the quantitative 

estimation of certain groups in phospholipins 403 

E. V. McCoLLUM and Marguerite Davis: The influence of the 
^ plane of protein intake on growth 415 



Contents v 

P. A. Leyenk and F. B. La Forc3e: On the mutarotation of phenyl- 

osazones of pentoses and hexoses 429 

P. A. Levene and F. B. La Forge: On chondroitin sulphuric acid. 

IV 433 

Joel A. Sperrt and Leo F. Rettger: The behavior of bacteria 

towards purified animal and vegetable proteins 445 

Robert M. Chapin and Wilmer C. Powick: Correction. An im- 
proved method for the estimation of inorganic phosphoric acid 

in certain tissues and food products 461 

R. J. Anderson: Concerning the organic phosphoric acid compound 
of wheat bran. IV. The occurrence of inosite triphosphate 

in wheat bran 463 

R. J. Anderson: The hydrolysis of phytin by the enzyme phytase 

contained in wheat bran 475 

R. J. Anderson : The hydrolysis of the organic phosphorus com- 
pound of wheat bran by the enzyme phytase 483 

R. J. Anderson : Concerning phytin in wheat bran 493 

£. C. Kendall: A method for the decomposition of the proteins of 

the thyroid, with a description of certain constituents 501 

G. E. CuLLEN and A. W. M. Ellis: The urea content of himian 

spinal fluid and blood 511 

Walter A. Jacobs and Michael Heidelberoer: Mercury deriva- 
tives of aromatic amines. I. Contribution to the structure of 
primary and secondary p-aminophenylmercuric compounds... 513 

P. A. Levene, C. J. West, and J. van der Scheer: The preparation 

and melting points of the higher aliphatic hydrocarbons 521 

Mary B. W ishart: Animal calorimetry. IX. The influence of meat 

ingestion on the amino-acid content of blood and muscle 535 

Frank A. Csonka: Animal calorimetry. X. The rate at which 

ingested glycocoll and alanine are metabolized 539 

Graham Lusk (with the technical assistance of J. A. Riche): Animal 
calorimetry. XI. An investigation into the causes of the 
specific dynamic action of the foodstuffs 555 

Stanley R. Benedict and Ethel H. Hitchcock: On the colorimet- 

ric estimation of uric acid in urine 619 

Stanley R. Benedict: On the colorimetric determination of uric 

acid in blood 629 

Stanley R. Benedict: Studies in uric acid metabolism. I. On the 

uric acid in ox and in chicken blood 633 

E. V. McCollum and Marguerite Davis: Nutrition with purified 

food substances 641 

Walter A. Jacobs and Michael Heidelberoer: The quaternary 
salts of hexamethylenetetramine. I. Substituted benzyl hal- 
ides and the hexamethylenetetraminium salts derived therefrom. 659 

\N alter a. Jacobs and Michael Heidelberoer: The quaternary 
salts of hexamethylenetetramine. II. Monohalogcnacctyl- 
benzylamines and their hexamethylenetetraminium salts 685 



vi Contents 

Frederic Fencer: On the presence of iodine in the human fetal thy- 
roid gland 605 

Herbert H. Bunzel: On alfalfa laccase 097 

Alfred W. Bosworth : Human milk 707 

Helman Rosenthal and P. F. Trowbridge: The estimation of fat. 711 

Index to Volume XX 719 

Proceedings of the American Society of Biological Chemists vii 




PROCEEDINGS OF THE AMERICAN SOCIETY OF 

BIOLOGICAL CHEMISTS. 

Ninth Annual Meeting. 
St. Louis, Mo., December 28-30, 1914. 



Vll 



PROCEEDINGS OF THE AMERICAN SOCIETY OF 

BIOLOGICAL COBMISTS. 

PRESIDENTIAL ADDRESS. 

THE INFLUENCE OF FOOD ON METABOLISM. 

By graham LUSK. 

In 1881 Voit laid down the principle that the intensity of metab- 
olism in the cells was modified by the quality and quantity of 
the food materials brought to them by the blood. He believed 
that the inherent power of the cells to metabolize was augmented 
by the presence of increased quantities of foodstuffs. Rubner 
developed another conception. He declared that the funda- 
mental metabolism of a normal warm-blooded animal was always 
constant and that the effect of food ingestion did not change this. 
The increased heat production which followed the taking of food 
was due to heat developed from a lot of intermediary reactions 
and oxidations and had nothing whatever to do with the funda- 
mental level of the cellular requirement of energy which was 
entirely unchanged. Thus, when protein was metabolized 
it could supply energy for the maintenance of true cellular activity 
in so far as glucose was produced from it, whereas other inter- 
mediary cleavage products were simply oxidized with the pro- 
duction of extra heat, which was in no way involved in the life 
processes of the cells. The utilization of energy in protein might 
be compared with the burning of a tree as fuel for the steam 
engine, the trunk of the tree being used as fuel within the engine 
for the production of power, whereas the limbs and- twigs are 
burned as brush outside and supply only heat. 

In 1907 when I saw Voit just a few months before his death, 
in the seventy-seventh year of his age, he expressed his strong 
dissent from Rubncr's views. At that time Rubner's theory 
had achieved wide acceptance, and the then known facts seemed 
to accord with his arguments. 

••• 



Society of Biological Chemists ix 

The theory of Zuntz that the specific dynamic action of the 
foodstuffs is chiefly due to intestinal activity no longer merits 
serious discussion. 

If one gives meat in large quantity to a dog, the heat production 
may be nearly doubled. During the second hour it reaches al- 
most its maximal height. The third hour registers the highest 
metabolism, and the heat production remains at this level for 
nearly twelve hours, during which the nitrogen in the urine 
manifests a high and almost even output from hour to hour. 
The metabolism then gradually falls and nearly reaches the 
basal level about the twenty-first hour. During the hours of the 
high metabolism the heat value of the protein destroyed as cal- 
culated from the nitrogen in the urine is far in excess of the heat 
eliminated by the dog. This is due to the fact that part of the 
carbon-containing radicles derived from the protein destroyed is 
retained in the organism. This retained carbon might be depos- 
ited as fat or as glucose. That it is retained in the form of glucosje 
is determined on the basis of the oxygen consumption and the 
computed heat value of the material retained. If it be assumed 
that the carbon is retained as glucose, the quantity of oxygen 
absorbed fits with the theory. If it were retained as fat, 10 
per cent less oxygen would have been required than was actually 
found. Calculated on the basis that the carbon is retained as 
glucose, indirect and direct calorimetry agree closely. Twenty per 
cent of the energy'- value of protein was found to be thus capable 
of retention in the form of glucose. Erwin Voit at one time 
put forward the theory that the specific dynamic action of 'pro- 
tein was due to the conversion of protein into fat. The above 
experiment disproves this idea. It will be shown later that the 
conversion of glucose into fat involves little energy change, so 
that even though over-ingestion of protein should be pushed so 
far as to cause the s>Tithetic production of fat from protein, this 
event would not noticeably affect the heat production. 

Experiments were instituted with the intention of more fully 
establishing the truth of Rubner's theories of specific dynamic 
action. It was known that glycocoU and alanine were completely 
convertible into glucose in the diabetic organism, whereas glu- 
tamic acid was in part so converted, three of its five carbon atoms 
passing into glucose, the other two being oxidized. It follows 



X Presidential Address 

from Rubner's hypothesis that glycocoll and alanine should exert 
no specific dynamic action, whereas glutamic acid should mani- 
fest this phenomenon. The reverse proved to be true; glycocoll 
and alanine are capable of greatly increasing the heat production, 
whereas the strong di-basic glutamic acid is without influence. 
Glycocoll and alanine produce powerful effects lasting eight and 
five hours respectively, whereas on giving those quantities of 
glucose into which the amino-acids are convertible an almost 
negligible influence is observed. 

The increase in metabolism after giving glycocoll and alanine 
together is equal to the sum of the effects produced by either 
alone. Furthermore, the increase of metabolism after giving 
20 grams of glycocoll is twice as great as after giving 10 grams. 
Similar relations obtain after giving different quantities of alanine. 
This accords with Rubner's discovery that the intensity of the 
specific dynamic action is proportional to the quantity of pro- 
tein ingested. When one compares the heat-increasing power of 
glycocoll and alanine upon metabolism, it is found that this 
power is not proportional to their respective abilities to form 
sugar, but rather to the number of molecules of glycoUic and 
lactic acids which they are respectively supposed to yield on 
deanaination. 

It was found in one experiment that the entire energy content 
of the ingested glycocoll reappeared in the extra output of energy 
given off by the dog in the form of heat. The course of inquiry 
into this phenomenon which naturally suggests itself is whether 
glycdcoll is without action upon the body cells; that is, whether 
it merely explodes and yields heat, or whether it directly stimu- 
lates the cells thereby raising metabolism to a higher level. This 
point was determined by giving glycocoll to a phlorhizinized 
animal. Under these circumstances there is no oxidation of 
the material ingested and the energy content of the glycocoll is 
eliminated in the urine in the form of sugar and urea. The 
metabolism was largely increased, notwithstanding the fact that 
there was no oxidation of the ingested material. Exactly the 
same phenomenon followed the ingestion of alanine in phlorhizin 
glycosuria. The ingestion of glucose was without effect even aft^r 
giving 70 grams. The cause of the specific dynamic action of 
glycocoll and alanine therefore lies in a chemical stimulation of 



Society of Biological Chemists xi 

the cells causing them to metabolize more material. This confirms 
the older view of Voit that the action of food increases the power 
of the cells to metabolize. 

The chemical stimulus to the cells does not reside in the amino- 
acids themselves, for (1) when meat is given and amino-acids 
are retained in the body for the synthesis of new protein there 
is no specific djrnamic action, (2) there is also no accumu- 
lation of amino-acids in the tissues under these circumstances, 
and (3) the hours of high metabolism after giving glycocoll and 
alanine are the hours of the maximal metabolism of these amino- 
acids. From such facts it is obvious that metabolism products 
of amino-acids like glycoUic and lactic acids are indicated as the 
substances which are the chemical stimuli. One recalls in this 
connection the permanently increased metabolism in phosphorus 
poisoning, in severe anemias, and in persons living at high alti- 
tudes, under all of which conditions lactic acid is found in in- 
creased amounts in the blood and often in the urine. 

That the chemical stimulus acts on protoplasm directly and 
not through excitation of the nervous system is to be inferred 
from the experiments of O. Frank and F. Voit who noticed a 
large increase in the heat production of curarized dogs after giv- 
ing them meat. 

External cold acts reflexly through the nervous system to 
increase metabolism in a fasting animal and thus prevents a fall 
in body temperature^ Rubner has called this the ''chemical 
regulation*' of body temperature. According to his hypothesis 
the "free heat'' liberated in the intermediary metabolism of 
protein can be used in lieu of that derived from the increased 
metabolism induced through the effect of cold. In the light 
of the newer researches, however, the extra heat necessary to 
preserve the body from a fall in temperature may be directly 
derived from an increased metabolism of the cell itself, whether 
this be induced by nerve action or by direct chemical stimulation. 

It has been found that ingested leucine and tyrosine also in- 
crease metabolism, though to a lesser extent than do glycocoll 
and alanine. 

It may be that the mass action of the various fragments pro- 
duced in the breakdown of protein in metabolism is also a con- 
tributar>' factor in the higher production of heat, hut. that it is the 



xii Presidential Address 

main factor is negatived by contrasting the different effect of 
20 grams of glutamic acid with that of 20 grams of glycocoll, 
t)^e effect of the first being nil and that of the latter powerful. 

Passing now to the consideration of the metabolism as influ- 
enced by the ingestion of fat and carbohydrates, it is found that 
an increase occurs whenever either fat or sugar is giveckj Fat 
pours into the blood stream through the thoracic duct, and ac- 
companying this increase in the food supply to the cells the metab- 
olism rises. The behavior of sugars has especially been studied. 
The ingestion of 50 grams of glucose by a dog causes at first an 
increase in the percentage quantity of blood sugar, the metabolism 
rises and is maintained at a high level during the three or four 
hours usually required for the absorption of the material. At the 
end of the second hour the blood sugar has fallen to the normal 
percentage quantity, but the blood has become much more dilute. 
Little urine is secreted until the last hour of the high metabolism; 
during that hour a large volume of urine is eliminated, and dur- 
ing the following hour the original basal metabolism is reached. 
During the period of high metabolism the respiratory quotient 
approximates unity, whereas subsequently this quotient may fall . 
to a level which indicates the oxidation of a mixture of carbo- 
hydrate and fat. With the cessation of absorption the influx 
of glucose molecules which were carried by an increased volume 
of blood ceases, their effect on the cells is cut off, and the metab- 
olism falls. After giving 70 grams of glucose the heat production 
may be 35 per cent above the basal metabolism during the hours 
of increased metabolism. All these details are obscured in long 
experiments lasting twenty-four hours, fit seems as though the ] 
quantity of fuel which nourished the cells determined the height 
of metabolism in the older sense of Voit. That the change in 
metabolism is not due to altered osmotic relations is evidenced 
by the fact that solutions of urea or common salt when ingested 
by a dog have no influence on his metabolism. — -:]) 

For a more complete understanding of the probable action of 
carbohydrates one must consider their intermediary metabolism. 

Dakin has shown how methyl glyoxal maybe obtained from 
both lactic acid and alanine; how it may also be formed in vitro 
from glucose, and be converted into glucose if it be given in phlor- 
hizin glycosuria. He gives the name glyoxalases to those tissue 
enzymes which rapidly convert methyl glyoxal into lactic acid. 



Society of Biological Chemists xiii 

The reactions which probably occur in the organism may be writ- 
ten as follows: 

CH, . CHOH . COOH \ 

Lactic acid \ CH,COCHOi=±C.Hi,0. 

/ Methyl 
CH, . CHNHj . COOH/ glyoxal Glucose 

Alanine 

The formation of methyl glyoxal as an intermediary product 
of carbohydrate metabolism explains the sjmthetic production 
of glucose from fructose in the organism. For if fructose be first 
converted into methyl glyoxal radicles which are free from asym- 
metric carbon atoms, the latter can then be synthesized into glu- 
cose, the sugar of the body. 

It has been afEmied by Woodyatt that lactic acid is not pro- 
duced from carbohydrate in the organism except in the cases of 
oxygen insufficiency. The first cleavage products of methyl 
glyoxal may then be acet aldehyde and formic acid. 

CHj CHj CHj 



2 CO + 2H^ -> 2 CHO -> CH, 

I I 

CHO 2HC00H CH, 



Methyl glyoxal Acetaldehyde COOH 

formic acid 

Butyric acid 

In the course of the usual metabolism of carbohydrate these 
substances would be oxidized. If, however, large quantities of 
carbohydrate entered the circulation, two molecules of acetalde- 
h^'de might be condensed to form butyric acid in accordance with 
the suggestion of Magnus-Levy. The building of higher fatty 
acids from butyric acid would require the addition of radicles 
consisting of — CHa— CH2— . Whatever may be the character 
of the intermediary changes, it is evident that the removal of 
CO2 from methyl glyoxal would leave the radicle — CH2 — CH2~ ► 

CHa 



C|0 

r 

Hi CO 



xiv Presidential Address 

The high respiratory quotient after giving carbohydrate in ex- 
cess is due to this intermediary liberation of carbon dioxide. 

The process of the conversion of carbohydrate into fat does 
not appear to be one involving any considerable increase in total 
.__ heat production. / Seventy grams of glucose were given to a dog 
on three different occasions. The extra carbon dioxide elim- 
ination which was not due to protein and carbohydrate com- 
bustion and which gave the high respiratory quotients always 
found after large carbohydrate ingestion, amounted to 0.8, 1.07, 
and 1.53 liters respectively, whereas the heat production as cal- 
culated amounted to 75.3, 75.7, and 75.6 calories. The height 
of the respiratory quotient is therefore not an index of increased 
metabolism. 

If one accepts Bleibtreu's formula as the simplest expression 
of the conversion of carbohydrate into fat, 

270.6 gm. glucose = 100 gm. fat + 115.45 gm. CO2 4- 54.6 gm. HjO 
997.2 calories = 950.0 calories 

the reaction is evidently exothermic, 4.7 per cent of the heat being 
liberated. If the heat evolved be measured on the basis of the 
extra carbon dioxide production, 1 liter of such carbon dioxide 
would have a value of 0.8 calories, or less than one-sixth the caloric 
equivalent of a liter of carbon dioxide obtained from the oxida- 
tion of glucose in the ordinary manner. 

This discussion makes it evident that when carbohydrate is 
converted into fat the heat production may be calculated from 
the oxygen absorption with a slight addition for the exothermic 
production of heat based on the elimination of carbon dioxide 
in excess of the requirement of a non-prot€fin respiratory quo- 
tient of unity. Such calculations show that direct and indirect 
calorimetry agree within 1 per cent. These experiments indicate 
that there is a definite upper limit for the metabolism of glucose 
molecules above which limit fat formation may occur from the 
assembled methyl glyoxal molecules which arc present in excess, 
and that this formation of fat is accompanied by only a very 
slight increase in the heat production. 

Benedict has stated that the cause of the specific dynamic 
action of carbohydrate is due to acid stimulation, partly basing 
his argument upon an experiment in which he gave 100 grams 



Society of Biological Chemists xv 

of fructose to a diabetic and witnessed an increase in metabolism 
of 30 per cent, although the fructose was not oxidized but was 
completely eliminated as glucose in the urine. It may be noted 
that this 30 per cent increase in heat production is very much 
greater than that which usually occurs in the human being after 
food ingestion. Thus, DuBois has found that the ingestion of 200 
grams of glucose by a normal man causes the metabolism to rise 
only 10 per cent. Furthermore, when 10 grams of fructose are 
given to a phlorhizinized dog there is no increase in the heat pro- 
duction, though the fructose is completely eliminated in the urine 
in the form of glucose. This contrasts with a large increase in 
metabolism when 12.5 grams of glycocoU, which are also con- 
vertible into 10 grams of glucose, are ingested in phlorhizin gly- 
cosuria. If it be permitted to assume that methyl glyoxal mole- 
cules are the intermediary products between fructose and glu- 
cose, then methyl glyoxal exerts no specific dynamic action in 
the sense of being a chemical stimulus of metabolism. The ex- 
periment performed by Benedict requires repetition and con- 
firmation. 

As regards the behavior of the different sugars in the normal 
organism, the work of Johanssoh and of Benedict has been con- 
firmed in the demonstration that fructose induces a greater heat 
production than glucose. Johansson gave the probable explana- 
tion: glucose was immediately ready for glycogen formation 
and fructose could not be immediately so deposited and was 
therefore oxidized in higher measure. This idea may now be 
amplified if one conceives that glucose may be at once removed 
and stored as glycogen, whereas fructose must first be entirely 
converted into the more readily oxidizable methyl glyoxal mole- 
cules, affording thereby an immediate and plentiful supply of 
small molecular fragments available either as food for the cells or 
as material for the construction of glucose or glycogen. 

It may be incidentally remarked that it was found that galac- 
tose was not readily oxidized by the dog, and that lactose had no 
effect whatever upon its metabolism, indicating the absence of 
lactase from the intestine. 

Experiments were instituted in which the effect of the inges- 
tion of 50 and 70 grams of glucose was compared with the effect 
of 50 grams of glucose to which 20 grams of glycocoll or of alanine 



xvi Presidential Address 

were added. Fifty grams of glucose increased the heat produc- 
tion 30 per cent, and 70 grams 35 per cent. There was little 
difference. Twenty grams of glycocoll increased it 36 per cent, 
and the same amount of alanine 32 per cent. Combined, 50 
grams of glucose and 20 grams of glycocoll are the glucose equiva- 
lent of 66 grams, and yet when they were given together the metab- 
olism increased 56 per cent. Glucose and alanine in similar 
quantities are a glucose equivalent of 70 grams and caused an 

increase of 53 per cent in the heat production ^' It is obvious that 

an increase in the quantity of glucose when this is given in large 
amounts, scarcely affects metabolism, whereas when the chemical 
stimulus from the metabolism products of the amino-acids acts 
on the cells in conjunction with a plentiful supply of glucose, the 
resultant effect is nearly equal to the sum of the two individual 
influences. This points to a distinct difference between the 
cause of the specific dynamic action of glucose and that of alanine, 
which latter is convertible into lactic acid and eventually into 
glucose. 

It will be recalled that Ward has shown in a mountaineering 
trip on Monte Rosa that the carbon dioxide tension in the alveoli 
falls as an accompaniment of the rising acid content of the body. 
If there were an acid production after carbohydrate ingestion 
one would expect to find the carbon dioxide content of the blood 
reduced from the normal, whereas unpublished experiments of 
Dr. A. L. Meyer show that this is not the case, the quantity 
remaining unchanged. One may accept this as added proof that 
the specific dynamic action of carbohydrate is not due to acid 
stimulation. 

Lactic acid from alanine or glycoUic acid from glycocoll may 
therefore raise the level of cell activity through direct stimula- 
tion, and if fragments of glucose metabolism be present in quantity 
these may enter as increased fuel to produce yet higher metab- 
olisin in the cells than the oxy acids would alone induce. _ 

Also, when alcohol is given with glucose the metabolism rises 
above the level it would have attained had glucose been admin- 
istered alone. The respiratory quotient falls and the cells oxi- 
dize both alcohol and the fragments of glucose metabolism, and 
produce almost as much extra heat as the sum of the quantities 
of heat which each material would have induced alone. 



Society of Biological Chemists xvii 

Of such nature is the metabolism of plethora. The influx i 
of carbohydrate of fat or of alcohol enables the cells to oxidize ' 
at a higher level through the increased mass action of food par- ; 
tides which are available. On the other hand, the metabolism 
products of glycocoU and alanine may directly stimulate pro- . 
toplasm without themselves being involved in the oxidative! 
process and this is called amino-acid stimulation. 

Finally, an analysis of the results obtained on diabetic pa- 
tients by Benedict and Joslin shows that the increase in metab- 
olism which has been reported as 15 per cent is only 5 or 10 
per cent. Benedict ascribes this increase to acidosis. In the 
phlorhizinized dog the metabolism may increase 70 per cent, in 
the depancreatized animal 40 per cent above the basal metab- 
ohsm. Yet depancreatized animals manifest scarcely any acidosis 
while the tissues of the human diabetic may be filled with the 
acetone bodies, as has been shown by Marriott. The increase in 
metabolism in dogs is better explained as being due to the' increased 
protein metabolism, as was first suggested by Rubner, and to the ' 
increased fat content of the blood; that is to say, to the dual 
mechanism of amino-acid stimulation and an existant plethora of 
food particles. 



ABSTRACT OF SCIENTIFIC PROCEEDINGS. 

THE EXCRETION OF CREATINE DURING A FAST. 

By F. D. ZE\UX and PAUL E. HOWE. 

(From the Biochemical Laboratories of Columbia University at Teachers 
College andjhe College of Physicians and Surgeons f New York,) 

Recent criticism^ of the results obtained with the Folin method 
for the determination of creatine in urine in the presence of 
acetone and aceto-acetic acid has thrown doubt upon the pres- 
ence of creatine in the urine of fasting man. We have deter- 
mined creatine in the urine of a fasting man throughout a seven 
day fast. The method of Graham and Poulton was employed 
for the removal of acetone and aceto-acetic acid and quantitative 
determinations were made of these substances together, and of 
/3-hydroxy butyric acid. Control experiments were made with 
untreated urine. Determinations before and after the appear- 
ance of the interfering substances showed the method to be 
accurate in their absence. Creatine was excreted on each fast- 
ing day in amounts comparable in most cases with those ob- 
tained in previous fasts under similar conditions. 

DETERMINATION OF CREATININE AND CREATINE ; THE OCCUR- 
RENCE OF CREATINE. 

By J. LUCIEX MORRIS. 

(From the Laboratory of Biological Chernistry,. Washington University y 

St. Ijouis.) 

Precipitation of creatinine as creatinine potassium picrate 
was made the basis of a method for determining creatinine and 
creatine. Such an estimation of creatine demonstrated the 
actual presence of this substance which is hydrolyzed into creati- 
nine, and precludes the possibility of confusion with other sub- 

* G. Graham and E. P. Poulton: Proc. Boy. Soc series B, Ixxxvii, p. 
205. 1914. 

xviii 



Society of Biological Chemists xix 

stances which are converted into forms giving the color reac- 
. tioD with picric acid and sodium hydroxide. The method 
served to separate the creatinine (excepting a practically con- 
stant fraction) from the interference of acetone bodies, and ' 
total creatinine from interference of frequently occurring unde- 
sirable products of hydrolyzing action (especially those arising 
from glucose). Thus separated, the double creatinine salt was 
brought into solution and the creatinine value determined as in 
the modified Folin method. Freedom from such interfering 
influences as these makes the method unquestionably worth 
while when their presence is suspected. 

Satisfactory application of the method was made in the analyses 
of nomial urines (plus creatine) with large amounts of sodium 
aceto-acetate and glucose added, diabetic urines, post-partum 
urines, and children's urines. In none of these did any inter- 
fering substance cause increase or decrease of the apparent values 
of either creatinine or creatine. In all classes of urines here 
mentioned the occurrence of creatine was demonstrated. 

THE INFLUENCE OF PROTEIN FEEDING ON THE ELIMINATION 

OF CREATINE IN STARVATION.' 

By WILLIAM C. ROSE. 

{From the Laboratories of Biological Chemistry of the University of PennsyU 
vania, Philadelphia^ and of the University of Texas , Galveston.) 

Experiments on dogs indicate that the feeding of diets rich 
in protein after complete extirpation of the pancreas does not 
cause the disappearance of urinary creatine. 

Contrary' to the observations of Cathcart, it has been found 
that in man the feeding of large amounts of protein causes a 
marked diminution in the creatine elimination during starvation. 
In severe cases of diabetes, however, protein feeding in man 
has no influence on the output of creatine. 

These results agree with those of Benedict and Osterberg on 
dogs with phlorhizin diabetes, and demonstrate that the dis- 
appearance of creatine from the urine following protein feeding 
in the normal fasting dog and in man is due to the carbohydrate 
arising from protein in metabolism. 

' Read by title. 



XX Scientific Proceedings 

THE NEPHELOMETRIC ESTIMATION OF PURINE BASES, INCLUD- 
ING URIC ACm, IN BLOOD AND URINE.' 

By SARA STOWELL GRAVES and PHILIP ADOLPH KOBER. 
(From the Harriman Research Laboratory , Roosevelt Hospital ^ New York.) 

I. Salskowski's reagent has been modified to meet nephel- 
ometric conditions, and it is shown that the modified reagent 
will precipitate xanthine, hypoxanthine, guanine, adenine, and 
uric acid in very dilute solutions (0.002 per cent) quantitatively. 

II. The use of a protective colloid has been introduced — clear 
solutions of egg albumin — for the purpose of keeping precipitates 
in suspension so that they may be estimated nephelometrically. 

III. It Is shown that an alkaline suspension of manganese 
dioxide, instead of an acid medium as has been used heretofore, 
will oxidize uric acid completely in three minutes, leaving the 
other purines unattacked. 

IV. It is shown that manganese dioxide is an excellent re- 
agent for the removal of alkaline sulphides from solution without 
the usual boiUng technique. 

V. It is shown that uric acid and other purine bases in urine 
may be quickly and fairly accurately estimated with the nephel- 
ometer. 

VI. It is shown that five volumes of 3 per cent sulphosalicylic 
acid is an excellent reagent for removing all coagulable protein 
from blood. By centrifuging for one or two minutes after the 
precipitation with sulphosalicyhc acid, the great bulk of protein 
can be removed, and if the supernatant liquid is shaken with a 
little talcum powder to cause agglutination of any remaining 
suspended protein, a perfectly clear filtrate can be obtained in 
five to ten minutes without boiling. 



* Read by title. 



Society of Biological Chemists xxi 

DEXTROSE CONTENT OF THE EGG OF THE COMMON FOWL.« 

By M. E. PENNINGTON, N. HENDRICKSON, E. L. CONNOLLY, 

AND B. M. HENDRIX. 

(From the Food Research Laboratory, Bureau of Chemistry , U. S. Depart- 

meni of Agriculturej Philadelphia.) 

The dextrose of the whites of fresh hens' eggs is found to vary 
from 0.3 to 0.6 of 1 per cent, averaging 0.44. The yolk contains 
from 0.08 to 0.24, averaging 0.08. During incubation the amount 
decreases in fertile, but is unaffected in non-fertile eggs. Com- 
mercial frozen egg contains from 0.28 to 0.38 of 1 per cent, aver- 
aging 0.34. The eggs rejected as unfit for food, except musty 
eg{^, contain much less than the good eggs. Storage of shell 
eggs for periods up to six months at a temperature close to 32° 
does not affect the dextrose content unless the eggs are infected 
by microorganisms. In conunercial dried egg the amount of 
dextrose present does not indicate the quality of the eggs from 
which the product was prepared, because a considerable part 
of the dextrose may be lost in the process of drying. In com- 
mercial frozen egg allowed to spoil, the dextrose disappears as 
decomposition progresses. Yolks decompose more readily than 
whites. A cleared solution in which dextrose can be determined 
is prepared by coagulating with heat and a little acetic acid, 
adding a large amount of washed alumina cream, and filtering. 

A METHOD FOR DETERMINING AND COMPARING THE LOCAL 
TOXICITY OF CHEMICAL COMPOUNDS. 

By H. J. CORPER. 

{From the Otho S. A. Sprague Memorial Institute and Pathological Labora- 
tory of the University of Chicago.) 

During the course of investigations on the pharmacological 
action of various copper amino-acid compounds it was found de- 
sirable to study the local toxicity of these compounds and to 
compare their action with that of copper salts of simple com- 
position; i.e. J copper sulphate. The method used in intracutane- 
ous tuberculin tests was admirably adapted for this purpose, 
requiring a I cc. tuberculin syringe graduated in 0.01 cc. and 

* Read bv title. 



xxii Scientific Proceedings 

fitted with a platinum needle (preferred for its resistance to the 
action of chemicals). Any suitable animal (preferably white and 
with the back shaved) is given intracutaneous injections of 0.2 
cc. of varying concentrations of the two compounds to be tested 
and compared (using, for instance, 1.0, 0.1, 0.01, and 0.001 per 
cent salt concentration), one compound being injected on one 
side of. the back and the other on the opposite side, equal con- 
centrations opposite each other, and noting from day to day the 
extent and type of the tissue changes produced. 

Thus it was noted that the local toxicity of a copper amino- 
acid mixture prepared from hydrolyzed egg albimain did not 
greatly differ from that of copper sulphate, the former pro- 
ducing, however, a hemorrhagic lesion while the latter pro- 
duced a simple necrosis. 

H. L. Huber continuing this investigation has found the same 
to be true of the leucinate and glutaminate of copper. 

DIGESTIVE PROCESSES IN LIMULUS.' 

By HELEN I. MATTILL and H. A. xMATTILL. 

(From the Marine Biological Laboratory, Woods Hole, and the Laboratory 
of Physiological Chemistry, University of Utah, Salt Lake City,) 

The alimentary tract comprises esophagus, stomach, intestine 
and end-gut; all are chitinous except the long straight intestine. 
A pyloric valve is invaginated into the intestine. From the 
intestine two pairs of ducts lead into the large digestive gland 
(*'liver") which is inextricably intertwined with the gonads; this 
mass fills the body cavity. The entire tract is alkaline. By 
means of a glass tube a red brown liquid is sometimes obtain- 
able from the stomach, which possesses the same digestive powers 
as the intestinal contents, a protease (leucine and tyrosine from 
fibrin), an amylase-maltase (dextrose from starch), and a lipase 
(Ijlmus-milk). Extracts of the intestinal mucosa and of the 
digestive gland showed the same digestive activity. The source 
of the secretion and the place of absorption are not yet certain. 
The intestine is filled with a clear, formless, jelly-like mass which 
completely invests and infiltrates the intestinal contents, and 

^ Read by title. 



Society of Biological Chemists xxiii 

which has perhaps the same function as the peritrophic mem- 
brane in insects, protecting the epithelium and allowing dis- 
solved material to pass out, and into the digestive gland. This 
jelly does not show carbohydrate on hydrolysis. Further work 
Is in progress and histological material is in preparation. 

EXPERIMENTS WITH dl-GLYCERIC ALDEHYDE. 

Bt R. T. WOODYATT. 

[Frnm the Olho S. A. Sprague Memorial Institute Laboratory of Clinical 

Research, Rush Medical College, Chicago.) 

Pure crystalline di-glyceric aldehyde made by a modified Wohl 
synthesis by E. J. Witzemann is capable of complete conversion 
into d-glucose in the fully diabetic organism, if given in low dilu- 
tions over a long time. If given in high concentration it fails 
to appear as glucose and apparently burns. By mouth doses of 
1.7 gm. per kilo of body weight caused diarrhea with unchanged 
triose in the stool and suppression of urine. Doses of 2.5 gm. 
per kilo subcutaneously are lethal, but cause no diarrhea and no 
excretion of unchanged triose by any route. 

THE LEVEL OF THE SUGAR IN THE BLOOD FLOWING FROM 
THE LIVER UNDER LABORATORY CONDITIONS. 

By J. J. R. MACLEOD and R. G. PEARCE. 

(From the Physiological Department^ Western Reserve Universityy 

Cleveland.) 

In order that the mobilization of sugar from the liver may be 
more thoroughly studied than has previously been the case, it 
is necessary to estimate the reducing power in small quanti- 
ties of blood taken at short intervals from the inferior vena cava 
opposite the hepatic veins. By a modification of Bang's micro 
method we have been able to do this with satisfactory accuracy 
in samples of blood measuring 2 cc. each and removed every 
five minut<?s. The results obtained so far are as follows: 

1. The experimental error of the method does not usually 
amount to more than 5 per cent. Occasionally it has been found 
to be greater. It is determined frequently for each experiment. 



xxiv Scientific Proceedings 

2. In a period of at least thirty minutes following the etheriza- 
tion and operative preparation of the animals, there was found to 
be a progressive fall in the sugar concentrartion. In observations 
on six dogs, this fall, expressed as a percentage of the original 
amount of sugar, amounted to 13, 14, 7.5, 10.5, and 23. The 
operations referred to consisted in introducing tracheal, carotid, 
and vena cava cannulae. 

3. After the first half hour, in four of the experiments there 
was a more or less distinct rise in sugar concentration, which 
was sufficient to bring it back to the original level in about one 
and one-half hours in three of the experiments. In the remain- 
ing two experiments the original decline persisted. 

4. It has been impossible to correlate these changes with any 
observable alteration in the physiological condition of the animals 
(arterial blood pressure, rectal temperature, respirations, depth 
of anesthesia). 

5. The extent of percentile rise occurring during a period of 
10 minutes varied as follows: 5.6, 4.5, 9.8, 6.5, 7, 11.5. These 
values are considerably below those previously obtained by us 
in observations made on blood similarly removed from animals 
in which the splanchnic or hepatic nerves were stimulated. 

A METHOD FOR THE DECOMPOSITION OF THE PROTEINS 
OF THE THYROID WITH A DESCRIPTION OF CERTAIN 

CONSTITUENTS. 

By E. C. KENDALL. 
{From the Mayo Clinicy Rochester, Minn.) 

The proteins of the thyroid may ha decomposed into simpler 
products by hydrolysis with 1 per cent sodium hydroxide in the 
presence of 90 per cent ethyl alcohol. The products so obtained 
may l)e separated into two groups by their solubility in acids. 
Those compounds precipitated by acids are designated Group A; 
those compounds soluble in acids are designated Group B. 

The iodine in the thyroid is evenly divided between these two 
groups, but there are many differences in the chemical properties 
of A and B iodine. Group A constituents are essentially acidic. 
Purification removes trv-ptophane, tyrosine, and lauric acid. 
The percentage of iotline is thus increased, and from these puri- 



Society of Biological Chemists xxv 

fied products I have isolated a crystalline compound containing 60 
per cent of iodine. Group B consists of amino-acid complexes, 
most of which are precipitated with 40 per cent ammonium sul- 
phate, but they are easily dialyzable. There is also a com- 
pound which reduces silver and mercury salts. I have named 
this compound R. 

The physiological activity of the thyroid is still retained in the 
constituents of A and B. In myxedema the dry, scaly skin is 
restored to moist normal condition by B. R, one of the con- 
stituents of B, reUeves muscle cramp. No toxic symptoms can 
be produced by any of the constituents of B. 

Marked toxic symptoms can be produced by A; namely, greatly 
increased pulse rate, loss of weight, nervousness, nausea, diarrhea, 
and severe headaches. The severity of these symptoms depends 
upon the iodine content of A. 

THE RELATION OF HYDROCHLORIC ACID TO THE MORPHOLOG- 
ICAL CHANGES INDUCED BY CHLOROFORM. 

By EVARTS a. GRAHAM. 

(From the Olho S. A. Sprague Memorial Institute Laboratory of Clinical 

Research J Rush Medical College, Chicago.) 

Central lobular liver necrosis of the type found in *1ate chloro- 
form poisoning" has been produced experimentally by other ali- 
phatic halogen substituted compounds; mz., CH2CI2, CCI4, CHBrs, 
CHIa, C2H5CI, CjHfiBr, C2H6I, C2H4Br2. It could not be pro- 
duced with ether or chloral hydrate. Chloroform in mtro is de- 
composed by oxygen into HCl and COCI2. The latter is further 
changed by water into 2HC1 and CO2. Analogous decomposi- 
tions occur with the other drugs named. The necrosis, and 
perhaps also the edema, hemorrhages, and fat infiltration which 
accompany it, are held to depend mainly on the action of the 
respective halogen acids formed in the breakdown of these sub- 
stances in the body. 

1. Oral or intraportal administration of HCl, in suitable con- 
centrations, produces liver necrosis, edema, hemorrhages, and 
fat accumulation. 

2. In chloroform poisoning central necrotic areas in the liver 
show a higher H+ and CI" content than control normal livers 



xxvi Scientific Proceedings 

when the H+ content is detennined by neutral red and naphthol 
blue, and the Cl~ content by AgNOa. 

3. The formation of the respective halogen acid after the 
administration of each of the drugs above mentioned is shown 
by the occurrence of the neutral salts of these acids in the urine. 

4. Alkali given in proper concentration inhibits the pro- 
duction of the necrosis and other changes by chloroform. 

5. The powers of CCI4, CHCI3, and CH2CI2 to produce these 
changes are proportional to the amounts of HCl which they 
may yield by decomposition; viz,, in the order mentioned. 

6. Of the ethyl compounds the chloride is least powerful in 
this respect and the iodide the most powerful. This agrees with 
their respective ability to form HCl, HBr, and HI outside of the 
body. 

7. The failure to produce the same type of changes with chloral 
hydrate agrees with its failure to form an appreciable amount of 
HCl in its breakdown in the body. 

VARIATIONS IN FACTORS ASSOCIATED WITH ACIDITY OF 

HUMAN URINE DURING A SEVEN DAY FAST AND 

DURING THE SUBSEQUENT NON-PROTEIN AND 

NORMAL FEEDING PERIODS. 

By F. D. ZEMAN, JEROME KOHX, and PAUL E. HOWE. 

{Frani the Biochemical Laboratories of Columbia University at Teachers 
College and the College of Physicians and Surgeons^ Sew York.) 

A study was made of the variations in acidity (true and titrat- 
ablc) of human urine with relation to the modifying factors pres- 
ent during fasting and recuperation. The range of variations 
of the acidity extended from a fairly acid urine, Ph 5.1 (third day 
of fast) to an alkaline urine, pg 8.0 (last day of the final period). 
The diet of the preliminary and final feeding periods was the same 
in nature, as that used in previous experiments.^ In the non- 
protein period cane-sugar, clarified butter, salts (alkaline mix- 
ture), and agar-agar were ing(»sted. 

Determinations were made of the H+ ion concentration (indi- 

* P. E. Howe, H. A. Mattill, and P. B. Hawk: Jour. Am. Chem. Sac., 
xxxiii, p. 568, 1911. Howe and Hawk: Proc. -Vm. Soc. Biol. Chemists, 2, p. 
65, 192, this Journal, xi, p. xxxi, 1912. 



Society of Biological Chemists xxvii 

cators); titratable acidity or alkalinity, using (a) phenol phthalein, 
(b) neutral red, (c) methyl orange; phosphates, anunonia, ace- 
tone, aceto-aeetic acid, and /8-hydroxy butyric acid. 

In the absence of exogenous phosphorus (fasting) we found the 
acidity (true and titratable), phosphates, acetone, aceto-acetic 
acid, and total nitrogen to vary together. During the non- 
protein post-fasting period we found an increased H"*" ion con- 
centration and acidity without an accompanying increase in the 
nitrogen excretion; acetone and aceto-acetic acid were absent. 
The increased excretion of ammofiia in fasting is correlated with 
that of /8-hydroxy butyric acid; when not influenced by this 
factor, as in the preliminary, non-protein, and final feeding 
periods, the anmionia excretion fluctuated with the H**" ion con- 
centration and the acidity. The low ammonia excretion in 
the final period showed the low H+ ion concentration and titrat- 
able acidity to result from a loss of fixed base. This phe- 
nomenon is apparently characteristic of recuperation (nitrogen 
retention). 

It seems probable that the increased nitrogen excretion during 
the early days of a fast of a human individual is i-elated to the 
meta}x)lic processes which result in the excretion of aceto-acetic 
acid. 

ON THE INFLUENCE OF SODIUM CARBONATE UPON GLYCO- 
SURIA, HYPERGLYCEMIA, AND THE^ RESPIRATORY METAB- 
OLISM OF DEPANCREATIZED DOGS. 

By B. KRAMER and J. R. MURLIX. 

(From the Physiological Laboratories of Cornell University Medical College ^ 
New York Cityf and of the University of loiva, Iowa City.) 

The present paper is a brief report of a large number of experi- 
ments on the influence of sodium carbonate administered in- 
travenously, subcutaneously, or orally on the fate of glucose in 
the depancreatized dog. The initial observation that 150 cc. 
of a 1 per cent solution of sodium carbonate greatly reduces 
the excretion of glucose in such an animal was reported a year and 
a half ago. To determine the fate of this retained glucose we 
naturally examined the blood first. Later, after greatly reducing 
the sugar in the urine, either by large doses given orally or by 



xxviii Scientific Proceedings 

smaller doses given intravenously, the respiratory exchange 
was studied before and after the administration of pure glucose 
by stomach. Finally, the liver and muscles have been examined 
for glycogen, the feces have been examined for sugar, and the 
urine has been tested for lactic acid. 

CONCLUSIONS. 

1. The administration of sodium carbonat€ to the depancrea- 
tized dog lowers the D : N ratfo as well as the sugar output per 
hour. 

2. That this is not due to a decrease in the permeability of the 
kidneys is shown by the fact that there is no increase and often 
a fall both in the percentage and absolute quantity of sugar 
in the blood. 

3. After the administration of carbonate and sugar a defi- 
nite retention of glucose occurs which is not due to its conversion 
into glycogen unless it be retained as glycogen in organs other 
than the liver or muscles. 

4. The sugar w^hich thus disappears from the urine is not ex- 
creted into the alimentary tract nor does it undergo condensation 
into substances that can be hydrolyzed by dilute acids. 

5. In the urines examined no lactic acid could be found even 
when the D : N ratio was minimal. 

6. For the time that the respiratory gases were studied the quo- 
tients obtained would indicate the combustion of but a minimal 
amount of glucose, if any. 

That glucoses may be burned in the diabetic dog after the re- 
peated administration of sodium carbonate is a probability, or 
that it may l)e converted into intermediary compounds other 
than lactic acid is another possibility, and these will furnish the 
subjects for future investigation. 



Society of Biological Chemists xxix 

NARCOTICS IN PHLORHIZIN DIABETES. 

By W. D. SANSUM and R. T. WOODYATT. 

(From the Otho S. A. Sprague Memorial Institute Laboratory of Clinical 

Research f Rush Medical College, Chicago.) 

The administration of phlorhizin alone in fasting does not re- 
move all glycogen from the body. This can be accomplished by 
subcutaneous injections of epinephrin (0.25 mgm.) at three hour 
intervals until the epinephrin no longer affects glycosuria. Any 
drug which in non-diabetic animals can cause hyperglycemia can 
also cause completely phlorhizinized, but not deglycogenized dogs 
to excrete *'extra sugar." 

Ether, nitrous oxide, acetaldehyde, etc., (narcotics) do this by 
virtue of their power to produce tissue asphyxia (increased acid). 
Among substances now regarded as capable of conversion into 
glucose a number have been tested only in incompletely phlor- 
hizinized dogs, and react like ether, etc. Pyruvic acid and acet- 
aldehyde are not sugar formers. 

ACCELERATION OF LIVER AUTOLYSIS. 

By H. C. BRADLEY. 
(From the Department of Physiology, University of Wisconsin, Madison.) 

Normal liver autolyzes till about 25 per cent of the nitrogen 
is not precipitated by tannic acid. The reaction is practically 
at equilibrium before ten days, and changes but slightly after 
three days. The addition of acid, or manganese salts leads to 
a more rapid and more complete hydrolysis. This increase is 
roughly proportional to the amount of acid or salt added until 
an optimum concentration is reached. In the case of HCl this 
optimum is about 1/50 mol.; in the case of MnCU it is about 1/10 
iiiol. At these concentrations the two reactions are nearly identi- 
cal. A similar increase of soluble nitrogen results from the 
addition of peptone, casein, and boiled liver. The addition of 
ovalbumin or edestin leads to no increase in soluble nitrogen. 
The former proteins constitute additions to the available sub- 
strate; the latter are not hydrolyzed at all by the enzymes and are 
not available in the reaction. The addition of the former gives 



XXX Scientific Proceedings 

an accelerated reaction which attains a higher soluble nitrogen 
level, and thus resembles the reaction in the presence of acid or 
MnCl2. 

The evidence appears to indicate that in the normal liver there 
are three protein groups or fractions. One represents the connec- 
tive tissue; another a mixture of hemoglobin and albumins; and 
the third a globulin type of protein. The connective tissue is 
insoluble and not attacked by the proteases under any of the 
conditions tried. It is, therefore, in no sense a substrate. It 
contains about 10 to 25 per cent of the total nitrogen of the 
liver. The second fraction, 45 to 65 per cent of the nitrogen, is 
not digested under normal conditions. It also is not to be con- 
sidered- substrate and has no mass effect in determining speed or 
equilibrium of the reaction. It may, however, be rendered avail- 
able by acids, salts of manganese, and other electrolytes, and 
when available acts as an added mass of substrate leading to 
more rapid digestion and a larger amount of soluble nitrogen. The 
third fraction, comprising from 25 to 30 per cent of the total nitro- 
gen, is readily available and normaUy is completely hydrolyzed 
under the experimental conditions in a relatively short time, — 
about two weeks. It is the mass of this fraction which deter- 
mines the speed and equilibrium of normal Uver autolysis. 

It is believed that most of the accelerations reported in the 
literature involve a shift upward of the equilibrium as well, and 
are only intelligible on the assumption that some of the non- 
available proteins are converted into substrate. In such con- 
ditions as phosphorus poisoning and long chloroform narcosis, 
the marked liver degenerations are due, first, to the toxic action 
on liver cells resulting in their death, and second, to the production 
of acid sufficient to neutraUze the alkalinity of the cells — which 
inhibits autolysis — and to convert in addition some of the non- 
available proteins into available. The result is a rapid and ex- 
tensive necrosis in vivOj while such tissue in vitro autolyzes faster 
and more completely than normal liver. 

There seems to be no evidence of an activation of a zymogen 
in these cases. 



Society of Biological Chemists xxxi 

ON THE NATURE OF THE HEPATIC FATTY INFILTRATION IN 
LATE PREGNANCY AND EARLY LACTATION. 

By V. H. MOTTRAM. 

During pregnancy sections of liver show an increase of fat 
under the microscope (Miotti). This increase is often very 
striking, occurs also in lactation; and simulates hunger and 
post-chloroform poisoning infiltrations. Cats and rabbits from 
the ordinary sources demonstrate this clearly. 

In an early experiment no less than 24 per cent of the fresh 
liver tissue was found to be pure non-volatile fatty acid with low 
iodine value. As the work developed the following data were 
collected: period of gestation; body and liver weights, percentage 
of pure non-volatile fatty acids of the Uver; iodine values of 
mesenteric hepatic and kidney fatty acids. 

Cats show too wide a variation from the mean in normal 
animals to yield unquestionable evidence in favor of infiltration. 
They require too much time and acconmiodation to bring their 
metabolism to an undisturbed level upon which a possible infil- 
tration might write itself. Rabbits are easier material. A 
month of quietude, separation, and ample dietary is enough to 
reduce individual variation within bounds. 

Control animals have a fatty acid percentage of fresh liver 
ranging from 2.36 to 3.19, with a mean of 2.62. Animals treated 
in the same way for the same time or longer, at full term or a 
day or two post-partum, show a range from 2.67 to 6.05, with a 
mean of 4.43. Taking other criteria, e.g., total liver fat or total 
liver fat divided by a function of the body weight, the result is 
the same: at or about the time of parturition there is an increase in the 
total fat in the liver. Histological results parallel this even in 
small percentage deviations. Argument from iodine values shows 
that the liver is infiltrated with fattv acid. 



xxxii Scientific Proceedings 

THE SYNTHESIS OF HIPPURIC ACID IN EXPERIMENTAL TAR- 
TRATE NEPHRITIS IN THE RABBIT. 

By F. B. KINGSBURY and E. T. BELL. 

{From the Departments of Physiology and Pathology of the Medical School 
of the University of Minnesota, Minneapolis.) 

Male rabbits, kept on a diet of carrots, were injected sub- 
cutaneously daily with sodium benzoate for a fore-period of sev- 
eral days, and then injected with from 0.3 to 0.7 of a gram of 
racemic tartaric acid dissolved in water and neutralized with 
sodium carbonate. This was injected into the muscles of the 
back. On the day that the tartrate was injected, the injection 
of sodium benzoate was postponed for two or three hours, for it 
was found that if it were injected at the same time as the tar- 
trate, it prevented the production of a severe nephritis. 

The urines were analyzed during the fore-period and during 
the period of nephritis, usually two or three days, for total nitro- 
gen, total benzoic, free Ixinzoic, and hippuric acids. The rabbit 
was then killed and an autopsy of its kidneys made. 

The degree of nephritis was shown by the phenolsulphone- 
phthalein test of Rowntree and Geraghty. A trace or zero of 
phthalein excretion is always associated with severe injury to 
the convoluted tubules, as shown by Potter and Bell. 

Severe nephritis of the convoluted tubules, as indicated by a 
zero or trace of phthalein excretion and confirmed by autopsy, 
did not affect materially the synthesis and excretion of hippuric 
acid. This is in agreement with the findings of Saloman and of 
Jaarsvcld and Stokvis, that in the rabbit hippuric acid is syn- 
thesized in other places as well as in the kidneys. 

ON THE RELATION OF THE OXYGEN TENSION OF THE AT- 
MOSPHERE TO C0B4BUSTI0N. 

By H. C. DALLWIG, A. C. KOLLS, and A. S. LOEVENHART. 

{From the Phannavnlogical Laboratory of the University of Wisconsin, 

Madison.) 

Since the time of Lavoisier much work has been done to show 
the similarities and differences between oxidation within and with- 
out the organism. In connection with our work on the effect 
of decreased oxygen tension in the respired air on the erythro- 



Society of Biological Chemists xxxiii 

cytes and hemoglobin, we exposed animals for one to two weeks 
to atmospheres of 10 per cent oxygen and 90 per cent nitrogen 
at the atmospheric pressure. We also exposed animals to atmos- 
pheres of similar low oxygen tension by partially evacuating the 
respirator^' chamber. We found that the physiological response 
is identical whether correspondingly low oxygen pressures are 
obtained by diluting the air with nitrogen or by partiaUy evacu- 
ating the chamber. We detennined to study also the effect 
of low oxygen tensions on the candle flame and the flames of 
other common combustible material. It was shown bv Clowes,^ 
working at atmospheric pressure, that various combustible 
gases and liquids would cease to bum at definite minimal con- 
centrations of oxygen. We have confirmed this. We have found 
that the paraffin as well as the ordinary candle is extinguished 
within a few seconds in an atmosphere of 15.5 per cent ()2 and 
84.5 per cent N2; viz.^ at an oxygen tension of 116.4 mm. Hg. 
We have found that by rapidly evacuating a vessel filled with 
air the flame is extinguished at a pressure of the residual air of 
95 mm. Hg., which corresponds to an oxygen tension of 19.8 
mm. Hg., or 2.6 per cent of an atmosphere. 

CONCLUSIONS. 

1. In certain combustions at least nitrogen is a powerful in- 
hibiting factor entirely apart from the proposition that it dilutes 
the oxygen. Thus the flame of the cahdle is just extinguished at 
1 16.4 nmi. Hg. oxygen pressure when the oxygen pressure is lowered 
by the addition of nitrogen, whereas the flame is just extin- 
guished at 19.8 nun. Hg. oxygen pressure if the lowering is effected 
by evacuation. In other words, the excess nitrogen retards the 
combustion enormously and in the presence of this gas in amounts 
required to make up the atmospheric pressure a partial pressure 
of oxygen six times as great as in the evacuated atmosphere 
has the same value in supporting combustion. 

2. The results bring out a striking difference between combus- 
tion and vital oxidation inasmuch as the latter depends purely 
on the oxygen tension, whereas in the former the nitrogen ten- 
sion plays an important role. 

^Clowes and Redwood: Detection and Estimation of Inflammable Gas 
and Vapor in the Air^ London, 1896. 



xxxiv Scientific Proceedings 

THE LEVEL OF BLOOD SUGAR IN THE DOG. 

By p. a. SHAFFER and R. S. HUBBARD. 

{From the Laboratory of Biological Chemistry^ Washitigton University, 

St. Louis.) 

Results already reported^ show that the normal level of blood 
sugar of the dog is about 0.05 per cent, but that if the animals 
are anesthetized with ether and the blood is drawn after an incision 
and dissection of a vessel, the amount of sugar found is usually 
between 0.10 per cent and 0.20 per cent, the values usually, but 
erroneously, accepted as normal. Later experiments appear to 
indicate that the higher results are in large part due to mild de- 
grees of asphyxia, even in those cases when the animal seems to 
be breathing well. If a tracheal cannula is inserted at once after 
the animal is anesthetized, and forced respiration is maintained, 
the blood sugar may rise sUghtly, but does not reach the values 
found without the forced respiration. 

The value of these observations would appear to be that by 
forced respiration the normal level of blood sugar of the dog 
may be almost, though not quite, maintained during operative 
procedures lasting even for several hours. 

It is suggested that forced respiration should be established 
in all experiments on dogs in which variations of the level of blood 
sugar are being observed as the result of any operative procedure. 

EXPERIMENTAL HYPERTHYROIDISM. » 
By VV. B. cannon, C. A. BINGER, and R. LITZ. 

FURTHER OBSERVATIONS ON THE ETIOLOGY OF GOITRE IN 

FISH." i« 

By DAVID AL4RINE. 

STUDIES ON EXPERIMENTAL CRETINISM.' 
By H. R. BASINGER and A. L. TATUM. 

« P. A. Shaffer: this Journal, xix, p. 297, 1914. 

• Presented from the Physiological Society in joint session of the 
Federation. 

" Presented from the Pathological Society in joint session of the 
Federation. 



Society of Biological Chemists xxxv 

A RESEARCH INTO THE FUNCTION OF THE THYROID.* '< 

By G. W. CHILE, F. W. HITCHINGS, and J. B. AUSTIN. 



THE EFFECT OF REPEATED INJECTIONS OF PITUITRIN ON 

MILK SECRETION.* '< 

By S. SLMPSON and R. L. HILL. 



THE ACTION OF PITXHTRIN ON THE MAMMARY GLAND.* 

By W. L. GAINES. 

ON THE MECHANISM OF PITUITOUS DIURESIS.* «> 
By F. B. KNOWLTON and A. C. SILVERMAN. 

THE SEVERAL FACTORS INVOLVED IN THE STANDARDIZATION 

OF PITUITARY EXTRACTS." 

By GEORGE B. ROTH. 

THE USE OF ALUMINUM HYDROXIDE IN CONNECTION WITH 
NITROGEN PARTITION IN URINARY ANALYSIS." 

Hy W. H. WELKER and GROVER TRACY. 

THE FECAL BACTERIA OUTPUT AS INFLUENCED BY DIETARY 

ALTERATIONS." 

By H. R. FISHBACH and P. B. HAWK. 

SOME INFLUENCES AFFECTING THE ACTION OF PHOSPHO- 

NUCLEASE." 

By OLAF BERGEIM. 

BIOLOGICAL OXIDIZABILITY AND CHEMICAL 

CONSTITUTION. II." 

By H. H. BUNZEL. 

" Presented from the Phannacological Society in joint session of the 
Federation. 

" Read by title. 



xxxvi Scientific Proceedings 

THE ACTION OF ALKALINE HYDROLYTIC AGENTS 

ON ALLANTOIN." 

By R. E. SWAIxV. 

CYANOGENESIS IN PLANTS. I. STUDIES ON SIEGLINGIA 

SESLEROIDES.'' 

By ARNO VIEHOEVER, C. O. JOHNS, and C. L. ALSBERG. 

SULPHUR PARTITION AS INFLUENCED BY WATER DRINKING.»^ 

By C. C. fowler and P. B. HAWK. 

THE INFLUENCE OF DEPANCREATIZATION UPON THE STATE 
OF GLYCEMIA AFTER INTRAVENOUS INJECTIONS OF DEX- 
TROSE IN DOGS." 

By I. ^. KLEINER and S. J. MELTZER. 

THE POSSIBILITY THAT SOME OF THE HEPATIC GLYCOGEN 
MAY BECOME CONVERTED INTO OTHER SUBSTANCES THAN 

DEXTROSE. 

By J. J. R. MACLEOD. 

ADRENAL DEFICIENCY." 
By R. S. HOSKINS. 

HYPOGLYCEMIA." 
By H. McGUIGAN. 

SOME EFFECTS OF ADRENALIN WHEN INJECTED INTO THE 

RESPIRATORY TRACT." 

By J. AUER and F. L. GATES. 

THE RELATION OF THE ADRENALS TO THE BRAIN."' »* 
By G. W. CRILE, F. W. HITCHINGS, and J. B. AUSTIN. 

" Read by title. 

" Presented from the Physiological Society in joint session of the 
Federation. 

" Presented from the Pharmacological Society in joint session of the 
Federation. 



Society of Biological Chemists xxxvii 

FURTHER OBSERVATIONS OF THE ORIGIN OF HYDROCHLORIC 

ACID IN THE STOMACH." 

By a. B. MACALLUM and J. B. COLLIP. 

THE EFFECT OF VARIOUS FLUIDS AND CEREALS ON GASTRIC 

SECRETION." 1' 

By C. C. fowler, M. E. REHFUS, and P. B. HAWK. 

■ 

THE RELATION OF THE DIGESTION CONTRACTIONS TO THE 
HUNGER CONTRACTIONS OF THE STOMACH (DOG, MAN).>^ 

By F. F. ROGERS and L. L. HARDT. 



RECUPERATION: NITROGEN METABOLISM OF A MAN WHEN 
INGESTING SUCCESSIVELY A NON-PROTEIN AND A NORMAL 

DIET AFTER A SEVEN DAY FAST.^' 

By F. D. ZEMAN, J. KOHX, and P. E. HOWE. 

THE DIASTASE OF THE BLOOD.»^ 
By H. McGUIGAX and C. L. V. HESS. 

THE RATE OF OXIDATION OF ENZYMES AND THEIR 
CORRESPONDING PRO-ENZYMES.>^ 

By W. E. BUHGE.* 

THE HARMFUL EFFECT OF AN EXCLUSIVE VEGETABLE DIET.^- 

By C. VOEGTLIN. 

THE EFFECT OF LONG CONTINUED FEEDING OF SAPONIN 
FROM THE BARK OF GUAIACUM OFFICINALE." '« 

By C. L. ALSBERG and C. S. SMITH. 



" Read by title. 

*' Presented from the Physiological Society in joint session of the 
Federation. 

** Presented from the Pharmacological Society in joint session of the 
Federation. 



xxxviii Scientific Proceedings 

FAT INFILTRATION OF THE LIVER AND KIDNEY INDUCED 

BY DIET." 

By E. L. OPIE and L. B. ALFORD. 

THE DETERMINATION OF BLOOD SUGAR (DEMONSTRATION). 

By P. A. SHAFFER. 

« 

THE EFFECT OF INTRAVENOUS INJECTIONS OF RADIUM ON 
THE URINARY NITROGEN AND SULPHUR PARTITION." 

By J. ROSENBLOOM. 

THE EFFECT OF EXTERNAL APPLICATION OF RADIUM ON THE 
METABOLISM OF A CANCER PATIENT.*^ 

By J. ROSENBLOOM. 

CARBOHYDRATE METABOLISM IN THE OYSTER.*^ 

By p. H. MITCHELL. 

STUDIES ON THE PATHOLOGY OF THE FEEBLE-MINDED. I. 
THE GLYCOSURIC REACTION AND ITS RELATION TO THEIR* 

PATHOLOGY." 

By AMOS W. PETERS. 

A METHOD FOR THE DETERMINATION OF BENZOIC ACID IN 

URINE." 

By G. W. RAIZISS axd H. DUBIN. 

THE SYNTHESIS OF HIPPURIC ACID IN THE ANIMAL BODY." 

By G. W. RAIZISS and H. DUBIN. 



" Presentod from the Pathological Society in joint session of the 
Federation. 

" Read by title. 



. 7 



VOL. XX JANUARY, 1915 No. 1 



THE JOURNAL 



OF ^■ 






. 'C •*, 



BIOLOGICAL CHEMISTRY 

VOUHDBD BT CWUCTIAM A. BBBTSB AND lUBrAIIfBO IN FAST BT TIB CNBIBTIAN A. BBBTBB 

MBllOBlAlt rUND 

EDITED BY 

H. D. DAKIN, New York City. LAFAYETTE B. IfENDEL, New Haven, Conn. 
B. K. DUNHAM, New York City. A. N. RICHARDS, Philadelphia, Pa. 

DONALD D. VAN SLYKE, New York City. 

WITH THB COLLABORATION OF 

J. J. ABBL, BifttmifB, Md. J. B. LBATHB3, TorQoto, CaOAda. 

R. H. CHITTBNDBN. New Hafta* Coaa. P. A. LEVEIfB. New York. 

OTTO FOUN, BostoB, Hast. JACQUBS LOEB. New Terk. 

WUXIAM J. GIBS. New Terk. A. S. LOBVBNHART, Hadiaea, Wla. 

L. J. HBNDXRSON, Cambridge, Kaia. GRAHAM LUSK. New Terk. 

RBID HUNT. BotCoa. Kaia. A. B. MACALLUM. Toronto, Caaada. 

W. A. JACOBS. New Yerk. J. J. R- MACLEOD, Cle?elaad, Ohio. 

WALTBR JONBS. Biltlmere. Md. JOHN A MANDBL, New York. 

J. H. KASTLB, Lexiagtoa, Ky. A. P. MATHBWS, Chkage, Dl. 

F. O. NOVY, Aaa Arber, Mkh. 

THOMAS B. OSBORNB, New Hatea, Ceaa. 

T. BRAILSFORD ROBERTSON, Berkeley, Cal. 

P. A. SHAFFER. St. Lonla, Me. 

A. B. TAYLOR. PhiUdelphia, Pa. 

F. P. UNDBRHILL, New Hatea, Coaa. 

▼. C. VAUGHAN, Aaa Arber, Mich. 

ALFRBD J. WAKBMAN, New HaToa, Coaa. 



PUBUSHBD MONTHLY 
BY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH 

FOR THE JOURNAL OF BIOLOGICAL CHElynSTRY, INC. 

2419-21 GREENMOUNT AVENUE, BALTIMORE, MD. 

Satered as aeeoad-elaaa matter. Auguat 1, 1911, at the Peat Oflice at Baltimore, Md., under the 

Aet of Mareh S. 1879. 



COPTBIOHT 1915 
BY 

The Joubnal of Bioloqical Chbuistbt 



STUDIES IN THE BLOOD RELATIONSHIP OF ANIMALS 
AS DISPLAYED IN THE COMPOSITION OF 
THE SERUM PROTEINS. 

m. A COMPARISON OF THE SERA OF THE HEN, TURKEY, DUCK, 
AND GOOSE WITH RESPECT TO THEIR CONTENT 

OF VARIOUS PROTEINS. 

By W. B. THOMPSON. 

(From the Rudolph SpreckeU Physiological Laboratory of the University of 

California.) 

(Received for publication, October 28, 1914.) 

The following investigations were undertaken at the suggestion 
of Dr. T. Brailsford Robertson, and constitute a third paper m a 
series of ' 'Studies in the Blood Relationship of Animals as Dis- 
played in the Composition of the Serum Proteins. "^^^ In this 
work I have employed Robertson's improved refractomctric 
method* for the quantitative determination of the various.proteins 
contained in the sera. 

In each instance the fowls were kept fasting for from eighteen 
to twenty hours previous to bleeding, but had free access to water. 
The sera were obtained by immediately defibrinating the fresh 
blood by shaking with glass beads. This must be done promptly, 
as the coagulation time for the blood of hens and ducks is about 
thirty seconds, for turkey blood about forty seconds, and for go<3se 
blood only about twenty seconds. Because of the small amount 
of blood yielded by each one of the smaller fowls, it was necessary 
to bleed several individuals for a sample of scrum, as noted below 
before each table of results. The analytical results reported are, 
in each instance, the average of at least two closely agreeing deter- 
minations made upon the same sample of serum. 

I T. B. Robertson: this Journal ^ xiii, p. 325, 1912. 

* J. Homer Woolsey: this Journaly xiv, p. 433, 1913. 

* Robertson: loc, cit. 



TVS JOUBMAL Of KOLOaiCAL CHEitimTRY, VOL. XX, NO. 1 



Proteins of Blood Sera 



A. Hen serum. 

The following are the results obtained from eight samples of 
sera from hens of between one and two years of age, each sample 
being from six hens. 

TABLE I. 

Hen serum. 



BZPERIIICNT 



" INSOLUBLE " 
GLOBULIN 



TOTAL GLOBULINS 



TOTAL ALBUMINS 



per caU 



■per cent 



1 

2 
3 



0.73 
0.73 
0.77 



0.04 
0.04 
0.04 



4 


0.73 =*= 0.04 


5 


0.73 =t 0.04 


6 


0.77 ± 0.04 


7 


0.73 =*= 0.04 


8 


0.73 =»= 0.04 



1.12 
1.11 
1.18 
1.18 
1.12 
1.18 
1.12 
1.12 



:±: 
db 



0.15 
0.15 
0.15 
0.15 
0.15 
0.15 
0.15 
0.15 



Average j 0.74 ± 0.04 1.14 * 0.15 



per cent 



TOTAL PBOTBINS 



per cent 



3.8 
4.0 
3.9 
3.9 
3.8 
3.9 
3.8 
3.8 



zt: 
dts 



0.2 
0.2 
0.2 
0.2 
0.2 
0.2 
0.2 
0.2 



5.0 
5.2 
5.0 
5.0 
5.0 
5.0 
5.0 
5.0 



db 
zt: 
=ir 
db 
zt: 



0.2 
0.2 
0.2 
0.2 
0.2 
0.2 
0.2 
0.2 



3.9 =*=0.2 



5.0 =fc0.2 



In each column the percentages show the per cent of that protein 
for each sample of serum. The figure following the =*= sign is, in 
each instance, the possible error due to a possible error of one 
minute in reading the angle of total reflection. 

Expressing each of the above proteins in terms of the percentage 
of the total protein content of the serum, the following figures 
are obtained: 



"Insoluble** globulin 14.9 

Total globulins 23,0 



Total albumins 77.0 



per cent 

ri5.3\ 

114.3/ 
f23.0\ 
\22.0J 

r78.0\ 
\77.0/ 



(=fc0.4) 
(*2.0) 
(=-=2.0) 



The first figure following each protein group represents the 
average percentage; the upper figure immediately following is the 
highest percentage observ^ed in any sample of the sera; the lower 
figure, the lowest percentage observed in any sample of the sera; 
and the figure in parenthesis, the plus or minus error in the esti- 
mate of these percentages that would be brought about by a possi- 
ble error of one minute in reading the total angle of reflection. 



W. B. Thompson 3 

W. D. Halliburton* estimates the proteins of hen serum to be 
as follows: 

per cent 

Total globulins 2.90 

Total albumins 1 .21 

Total proteins 4 .14 

That these figures differ in so great a degree from my results may 
be accounted for by the fact that Halliburton employed Ham- 
marsten's method of estimating the total globulins. Moreover, 
the birds employed by Halliburton were probably not fasted.^ 

B. Rooster serum. 



The following are the results obtained from determinations made 
upK>n four samples of serum from roosters aged four years and over, 
each sample being from three roosters. 





■ 


TABLE II. 










Rooster serum. 




EXPERIMENT 


** INSOLUBLE •• 
GLOBULIN 


TOTAL GLOBULINS 


TOTAL ALBUMINS 

per cent 


TOTAL PROTEINS 




per cent 


per cent 


per cent 


9 


0.70 =*= 0.04 


1.12 =fc 0.15 


4.9 =fc0.2 


6.0 =*=o.2 


10 


0.63 ± 0.04 


1.14 =*= 0.15 


5.1 ± 0.2 


6.2 ± 0.2 


11 


0.70 ± 0.04 


1.13 =*= 0.15 


4.7 ± 0.2 


5.8 ± 0.2 


12 


0.70 =*= 0.04 


1.12 =fc0.15 


4.9 =^0.2 


6.0 =*= 0.2 


Average 


0.60 =fc 0.04 


1.127 =*= 0.15 


4.9 =fc0.2 


6.0 ± 0.2 



Expressing each of the above proteins in terms of the percentage 
of the total protein content of the serum, the following relation- 
ship is shown: 



"Insoluble*' globulin 11.4 

Total globulins 19.0 

Total albumins 81 .0 



per cent 
12.0' 

10.7^ 
20.0 
18.0. 
82 .0' 
81.0. 



(^0.4) 
(*2.0) 
(=*=2.0) 



*W. D. Halliburton: Jour. Physiol, vii, p. 319, 18S6. 
* Compare the succeeding article, p. 7. 



Proteins of Blood Sera 



This table shows a marked difference in the total protein con* 
tent from that of the hen senun, the difference being chiefly in the 
amount of albumins present. It has been reported* for human 
serum that the organic content, of which over 90 per cent is pro- 
tein, is greater in the male than in the female. Unfortunately it 
was not possible to make determinations upon the sera of fowls of 
the same age of opposite sexes, and upon fowls of the same sex with 
as great a difference in ages as exists between the ages of the hens 
and roosters used. Hence I am unable to state whether this dif- 
ference in the proteins may be attributed to the influence of age 
or sex. 

C Turkey serum. 

The following results were obtained from determinations made 
upon separate samples of sera from three turkeys. 

TABLE m. 
Turkey serum. 



EXPERIMENT 



*• INSOLUBLE " 
GLOBULIN 



13 
14 
15 



Average 



per cerU 

0.47 =*= 0.04 
0.60 =*=0.04 
0.47 * 0.04 



0.48 -=0.04 



I 

TOTAL GLOBULINS ' TOTAL ALBUMINS 



per cent 

1.02 * 0.15 
0.98 =*= 0.15 
1.00 =*«0.15 



1.00=*= 0.15 



per cent 

5.2 =b0.2 
5.4 =fc 0.2 
5.0 =*=0.2 



6.23 =*=0.2 



TOTAL PROTEINS 

per cent 

6.3 i*=0.2 

6.4 =*= 0.2 
6.0=*= 0.2 



6.2=*= 0.2 



Expressing each of the above in terms of the percentage of the 
total protein content of the serum, the following figures are ob- 
tained: 

per cent 

"Insoluble" globulin 7.4 | ^gj (±0.4) 

Total globulins 16.0 {Jsoj (=*=2.0) 

Total albumins 84.0 (^oj ^^^'^^ 

D. Duck serum. 

The following results were obtained from determinations made 
upon duck sera. The number of ducks that were bled to obtain 

•O. Hammarsten: A Text Book of Physiological Chemistry, translated 
by Mandel, 6th edition, New York, 1911, p. 315. 



W. B. Thompson 



each sample of serum is noted in parenthesis after the experiment 
number. 

TABLE IV. 

Diick serum. 



WXrUMXMKWT 


"zksolubub" 

OLOBUUM 


TOTAL OLOBULXlfB 


TOTAL ALBUMOIB 


TOTAL PROTKXM8 




per Mill 


per cent 


per cent 


per cent 


16(1) 


1.41 *0.04 


1.90*0.15 


4.6*0.2 


6.5*0.2 


17 (1) 


1.56 *0.04 


2,01 * 0.15 


4.6*0.2 


6.6 *0.2 


18(2) 


1.53 *0.04 


2.00*0.15 


5.0*0.2 


7.0 * 0.2 


19(2) 


1.28*0.01 


1.72 *0.15 


5.0*0.2 


6.6 *0.2 


20(3) 


1.44*0.04 


1.91 *0.15 


4.8*0.2 
4.8*0.2 


6.7 * 0.2 


Average 


1.44 *0.04 


1.91 *0.15 


6.7 *0.2 


ExDress] 


inir each of th( 


5 above nrotei 


ns in terms of 


the t^tal nro- 



tein content of the serum, the following figures are obtained: 



"Insoluble" globulin 21.6 

Total globulins 26.0 



Total albumins 74.0 



pereerU 

r23.8\ 
\l9.l/ 

fai.ol 

126.0/ 
r74.0l 
\73.0/ 



(=*=0.4) 
(*2.0) 
(*2.0) 



E. Goose serum. 



The following results were obtained from determinations made 
upon separate samples from four geese. 



TABLE V. 
Goose serum. 



KXRBDnirr 



21 
22 
23 
24 



"xmboluble" 
globulin 



TOTAL GLOBULINS 



per cent 

0.73 * 0.04 
0.70*0.04 
0.73 * 0.04 
0.70 * 0.04 



Average 



0.71 * 0.04 



per cent 

1.11 * 0.15 
1.11 *0.15 

1.08 * 0.15 
1.08 * 0.15 



TOTAL ALBUMINB 



per cent 

3.0 * 0.2 
3.0 * 0.2 
3.0 * 0.2 
3.0 * 0.2 



TOTAL PROTBINS 



1.10 *0.15 



3.0 *0.2 



per cent 

4.2 *0.2 
4.2 * 0.2 
4.1 * 0.2 
4.1 * 0.2 



4.1 *0.2 



Expressing each of the above proteins in terms of the total pro- 
tein content of the serum, the following figures are obtained: 



Proteins of Blood Sera 



pet C€ni 



'^Insoluble" globulin 17 .2 

Total globulins 26.0 



Total albumins 74.0 



SUMMARY. 




(=*=0.4) 
(=*=2.0) 
(=*=2.0) 



The following table suimiiarizes the average results obtained 

TABLE VI. 



PEBCGNTAOE OF THE TOTAL PROTEINS IN THE SERA OF 



Hen 



^'Insoluble" globulin ' 14.9 

Total globulins 23.0 

Total albumins 77 .0 



Rooster i Turkey 



11.1 
19.0 
81.0 



7.4 , 
16.0 
84.0 1 



Duck 
21.6 

29.0 
72.0 



Groose 



17.2 
26.0 
74 



STUDIES IN THE BLOOD RELATIONSHIP OF ANIMALS 

AS DISPLAYED IN THE COMPOSITION OF 

THE SERUM PROTEINS. 

IV. A COMPARISON OF THE SERA OF THE PIGEON, ROOSTER, 

AND GUINEA FOWL WITH RESPECT TO THEIR CONTENT 

OF VARIOUS PROTEINS IN THE NORMAL AND 

IN THE FASTING CONDITION. 

By R. S. BRIGGS. 

{From the Rudolph Spreckels Physiological Laboratory of the University of 

Calif ornia.) 

(Received for publication, October 28, 1914.) 

In canying out the following series of observations I have made 
use of the refractometric method devised by Robertson for the 
determination of the relative amount of the various blood proteins.^ 

For each species of animal used in this investigation, observa- 
tions were taken in both the fasting and non-fasting condition, the 
latter being subsequently referred to as "normal." The length 
of the fast was in all cases at least twenty-four hours. 

The serum was obtained by cutting the jugular vein, defibrin- 
ating the blood with glass beads, centrifuging, and carefully re- 
moving the serum with a bulb-pipette. In the case of the pigeon 
and guinea fowl it was found necessary to use more than one bird 
in order to obtain a suflScient amount of serum for a complete 
examination. For example, in the first analysis with pigeon serum 
the amount obtained from four pigeons was found to be insufficient, 
and consequently six were used for each subsequent analysis. In 
the case of the guinea fowls, two were used for the first and three 
for the second analysis. In the case of the rooster, however, a 
sufficient amount could be obtained from one bird for a complete 
analysis. 

A. Pigeon serum. 

The following numbers indicate the percentages of the various 
proteins present in the original serum. 

^ This Journal, ziii, p. 325, 1913. 



8 



Proteins of Blood Sera 







TABLE I. 










Normal pigeons. 




QROUP NO. 


"insolublb" 
globulin 


TOTAl. 
GLOBULINS 


TOTAL 
ALBUMINS 


TOTAL 
PHOTKINS 




percent 


percent 


percent 


percent 


1 




0.93 =*= 0.15 


3.4^=0.2 


4.3 =fc0.2 


2 


0.33 =fc 0.04 


1.17 * 0.15 


2.6 =fc0.2 


3.8 =t 0.2 


3 


0.30 =fc0.04 


1.58=^0.15 


3.5=*= 0.2 


5.1 =*=0.2 


Average 


0.32 =*= 0.04 


1.22=^0.15 


3.1 =fc0.2 


4.4=*= 0.2 



The figure following the =*= sign is the possible error in the deter- 
mination due to a possible error of one minute in reading the angle 
of total reflection. 

Expressing each of the above mentioned proteins in terms of the 
percentage of the total proteins which they represent, the follow- 
ing figures are obtained: 



'ansoluble" globulin 7.3 

Total globulins 28.0 

Total albumins 72 .0 



percent 
{ 5.9} 

f 31 .0\ 
\22.0/ 

\68.o; 



(* 0.4) 
(* 2.0) 
(* 2.0) 



The first figure opposite each group represents the average per- 
centage; the upper figures immediately following, the highest 
percentage observed in any individual; the lower figure the lowest 
percentage observed in any individual; and the figure in paren- 
thesis, the plus or minus error in the estimation of these percent- 
ages which would be brought about by an error of one minute in 
reading the angle of total reflection. 

The following table indicates the results obtained with fasted 
pigeons. 







TABLE II. 








"INSOLUBLE" 
OLOBUUN 

percent 


Fasted 'pigeons 

TOTAL 
OLOBUUKB 

pfrcent 


• 

TOTAL 
ALBUMINS 




GROUP NO. 


TOTAL 
PBOTXINB 




per cent 


percent 


4 


0.13 =fc 0.04 


0.68 =fc0.15 


3.9 =fc 0.2 


4.6 =fc 0.2 


5 


0.13 ±0.04 


0.67 =*= 0.15 


4.3 =*= 0.2 


5.0 ±0.2 


Average 


0.13 =fc0.04 


0.67 =*=0.15 


4.1 =*= 0.2 


4.8 =*=0.2 



R. S. Briggs 9 

Expressing each of the above mentioned proteins in terms of 
the percentage of the total proteins which they represent, the fol- 
lowing figures are obtained: 



"Insoluble'' globulin 2.7 

Total globulins 14.0 

Total albumins 86 .0 



percent 

ri5.o\ 

\l3.0J 
/86.0\ 
VS5 .0 / 



(=*= 0.4) 
(=*= 2.0) 
(=*= 2.0) 



By comparison of Tables I and II, a decided increase of total 
albumins may be observed in the serum of the fasted animal. 
A similar result was obtained with rooster serum. 



B. Rooster serum. 

The following numbers indicate the percentage of the protein 
present in the original serum. 

TABLE m. 
Normal roosters. 



KOOSTBBNO. 


**IMBOLXniLB" 
OLOBUUN 


TOTAL 
OLOBXTUNB 


TOTAL 
ALBUKIia 


TOTAL 
PROTEINS 


1 
2 


perettU 

0.63^0.04 
0.66^0.04 


pereeni 

1.34 =*= 0.16 
1.16 ±0.16 


pereeni 

2.0=*= 0.2 
2.0*0.2 


pereeni 

3.3*0.2 
3.2*0.2 


Average 


0.54 -= 0.04 


1.25 *0.16 


2.0 =fc0.2 


3.3* 0.2 



Ebcpressing each of the above mentioned proteins in terms of 
the percentage of the total proteins which they represent, the fol- 
lowing figures are obtained: 



percent 



"Insoluble" globulin 16.4 

Total globulins 38 .0 

Total albumins 62.0 



\i6.i; 


(* 0.4) 


/ 41 .0\ 
136.0/ 


(* 2.0) 


res .01 

161 .0/ 


(* 2.0) 



With fasted roosters the following results were obtained: 



lO 



Proteins of Blood Sera 





"iNftOLUBlK" 
aiX)BUUN 


TABLE IV. 
' Fasted roosters. 




R008TKR NO. 


TOTAL 
GLOBULINS 


TOTAL 
ALBUMINS 

percent 


TOTAL 
PROTKINS 




per cent 


per cent 


per cent 


3 




1.56 =fc 0.15 


3.0=*= 0.2 


4.5 ±0.2 


4 


0.57 ± 0.04 


1.03 ±0.15 


3.1 =fc 0.2 


4.1 ± 0.2 


5 


0.49 =fc 0.04 


1.97 =fc0.15 


2.8 =*= 0.2 


4.7 ± 0.2 


6 


0.43 =fc0.04 
0.50 =fc0.04 


1.10 ± 0.15 


2.6 ±0.2 


3.7 ± 0.2 


Average 


1.41 =fc 0.15 


2.9 ±0.2 


4.13 ± 0.2 



Expressing each of the above mentioned proteins in terms of 
the percentage of the total proteins which they represent, the 
following figm'es are obtained: 

"Insoluble" globulin 12.1 

Total globulins 33 .0 

Total albumins 68 .0 




(=*=0.4) 
(* 2.0) 
{^ 2.0) 



By comparison of the percentages of total albumins listed in 
Tables III and IV, it will again be seen, as in the case of the 
pigeon, that the fasted animals show an increase in total proteins. 

Upon comparing these results with the data obtained by Thomp- 
son for fasted roosters,^ it will be seen that the content of total 
albumins which is indicated by his results considerably exceeds 
that obtained by mine. In both series of experiments the length 
of the fast was approximately the same, but the birds employed 
by Thompson in obtaining his data were much older than those 
used in obtaining the above results. The former were all over four 
years of age, while the latter were all between the ages of one 
and two years. This difference in age may be the source of the 
markedly higher figures for the total albumins found by Thompson. 

C. Guinea fowl serum. 

TABLE y. 
Normal guinea fowls. 



GROUP NO. 



"insoluble" 
globulin 



TOTAL 
GLOBULINS 



1 



percent 
0.23 ± 0.04 



per cent 
1.12 ± 0.15 



TOTAL 
ALBUMINS 



per cent 
2.6 ± 0.2 



TOTAL 
PROTEINS 



per cent 

3.7 ± 0.2 



* Compare the preceding article, p 1. 



R. S. Briggs 



II 



Lxpressing each of the above mentioned proteins in tenns of the 
percentage of the total proteins which they represent, the following 
figures are obtained: 

percent 

''Insoluble" globulin 6.2 =*= 0.4 

Total globulins 30.0 =*= 2.0 

Total albumins 70.0 =*= 2.0 



GBOUP NO. 



**in»oluble" 

OLOBUUN 



per cent 

0.20 ^ 0.04 



TABLE VI. 
Fasted guinea fowls. 



TOTAL 
OLOBUUKA 



TOTAL 
ALBUMINS 



per cent 

0.22 =fc 0.15 



per cent 

3.9 ^ 0.2 



TOTAL 
PBOTBIN8 



per cent 
4.1 ^ 0.2 



Expressing each of the above mentioned proteins in terms of the 
percentage of total proteins which they represent, the following 
figures are obtained: 

percent 

"Insoluble" globulin 4.7 =*= 0.4 

Total globulins 5 .0 =*= 2 .0 

Total albumins 95.0 =fc 2.0 

. These results, as in the case of the pigeon and rooster, again 
show a decided increase in the total albumins of the serum of the 
fasted birds. 



SUMMARY. 



The following table summarizes the average results obtained. 



Pigeon 



Rooster. 



Ctuinea fowl 



/normal 
\ fasting 
f normal 
\ fasting 
r normal 
\ fasting 



TABLE VU. 



"insoluble" 
globulin 



per cent 

7.3 
2.7 
16.4 
12.1 
6.2 
4.7 



TOTAL 
GLOBULINS 



per cent 

28 
14 
38 
33 
30 



TOTAL 
ALBUMINS 



per cent 

72 
86 
62 
68 
70 
95 



PRECIPITATION OF SERUM-ALBUMIN AND GLUTIN 

BY ALKALOIDAL REAGENTS. 

By PAUL J. HANZLIK. 

{From the LaborcUory of Physicochemical Biology, University of Vienna, 

Director: Prof, Wolfgang Pauli.) 

(Received for publication, November 1, 1914.) 

The study of the mechanism of the precipitation of proteins by 
so called alkaloidal reagents (tannin, iodine potassium iodide, po- 
tassium ferrocyanide, potassium mercuric iodide, tungstate, and 
pbosphomolybdate) has received practically no attention from the 
standpoint of physical chemistry, especially the isoelectric point. 
According to Michaelis,* the isoelectric point of serum is the point 
where the number of positive and negative charges of electricity 
held by the protein ions are about equal; that is, at this point 
there is a maximum of neutral particles of protein. This lies ap- 
proximately at the concentration of 2. 10-* of acid, such as ob- 
tains in a mixture of equal parts (t) of t^ acetic acid and -nr 
sodium acetate. Certain other phenomena which are exhibited by 
proteins at this point are, according to Pauli^ and his pupils, a 
minimum of viscosity as shown by serum and a maximum of pre- 
cipitation by ethyl alcohol. It would be interesting to know the 
behavior of other precipitants, such as the alkaloidal reagents, 
with respect to this point. It might also indicate whether the 
mechanism of precipitation by all alkaloidal reagents is the same 
or different. This has hitherto been regarded as identical. 

In the following experiments attempts have been made to as- 
certain where precipitation of serum and glutin takes place with 
respect to their isoelectric points, and how it is influenced by differ- 
ent degrees of acidity and the addition of salt. The conditions 
under which precipitation was observed have been expressed in a 
quantitative manner; that is, definite quantities of serum and 
glutin of known protein content were mixed with definite quanti- 

*L. Michaelis tfnd B. Mostunski: Biochem. Zisckr., xxiv, p. 79, 1910. 
» W. Pauli: Ztschr. f. Chem. u, Ind. d. KolL, xii, p. 222, 1913. 

13 



14 Precipitation of Serum- Albumin and Glutin 

■ 

ties of known strengths of acids, and to this the same quantitj' 
of the reagent of known strength was added. The total volume 
of the mixture was always the same. The effect of salt was studied 
in a similar manner. 

All experiments were performed at ordinary room temperature 
in test-tubes. The precipitates were observed when freshly 
formed and at the end of twenty-four hours. However, it is to 
be noted that, as a rule, the precipitates remained unchanged on 
standing, and the data have been compiled without reference to 
this factor. Precipitation was regarded to have taken place when 
the mixture upon the addition of the reagent became non-trans- 
parent. The different degrees of precipitation have been expressed 

TABLE !.• 
Degree of acidity represented by acetic-acetate ('*6ujfer") mixtures. 



Yff ACETIC ACID 
PROPORTION OF ACETIC ACID TO ACETATE IK MIXTURE: -jp 



TTT 



SODIUM ACETATE 



N 

T 



\ 



lH+1 
Ph 



1.44 X 10-« 
3.84 



0.9X10-«t 
4.05t 



4» 



IH + ] 
Ph 



\ 



i__._t ___ 

, 0.72 X 10-* 
j 4.14 

I 

!""■ i 



? 



i' 



! 



0.54 X 10-«t 
4.27t 



A 



0.36 X 10-* 
4.44 t 



0.27XlO-*t 1.8X10-* I 0.9X10* 



4.57' 



4.74 



5.05 



0.45 X KM 
5.35 



0.22 X 10-* 
5.66 



0.18 X 10-»t 
6.74t 



*In the tabloR, the varioua signs and abbreviatioDs used have meanings as follows: the plus 
sign (+) "■ precipitate present: the minus sign (— ) a no precipitate; si — slight; st ■■ strong; 
max » maximum; (7) ■■ precipitate doubtful; tr » trace. The order in which the solutions 
were fixed and the quantities used are appended as footnotes to each table. 

t Calculated. 

in ordinary terms as indicated in the various tables. The reagents 
were previously rendered neutral to litmus by the addition of 
sodium hydroxide, and the strengths of these were usually 5 per 
cent. The horse serum and glutin used in these experiments had 
been previously dialyzed for a period of five to six weeks, and were 
practically salt-free. The end concentration represented in the 
different experiments will be found appended to each table. The 
standard "buffer'' mixtures of acetic acid and sodium acetate were 
prepared according to Sorcnsen,' and the degrees of acidity that 
they represent are shown in Table I. In my work the concen- 



'Sorenson: Ergebn. d. PhymoL, xii, p. 393, 1912. 



Paul J. HanzHk 



15 



tration of sodium acetate in all experiments is always the same, 
only the variation in the acetic acid being present. The different 
concentrations of acetic and hydrochloric acids were prepared by 
dilution of standard solutions in the usual manner, and the con- 
centrations of these are indicated in the various tables. 

/ . Precipitation of serum and glutin in mixtures of acetic acid and 

sodium acetate. 

In mixtures of acetic acid and sodium acetate with proteins, 
the dissociated hydrogen ions are not bound by the protein, but 

TABLE II. 

Precipitation of serum and glutin by alkaloidal reagents in the 

acetic-acetate mixtures. 



N 

IT 
ACETIC Aao 



SODIUM 
ACXTATX 



i 



T 

i. 
I 



1 
1 

» 

I 

i 

Without acid 



lODINB 

POTA08IX7M 

IODIDE 



POTABSniM 
lOBBCXTBIC 
IODIDE, 
, 5 PER CENT 



SODIUM 

MOLTBDATE, 

6 PER CENT 



I SODIUU 
I PH06PHO- 
ITUNG8TATE. 
I 5 PER CENT 



s 



s 



G 1 8 



I 



I 



i 



+ 
+ 



+ 
+ 

+ 8l 



+ 8t 

+ 



+ 
+ 



+ 
+ 

4-sl 

+ 
+ ? 



+ 8t 

-|- less 8t 
-|- least 



+ 

max 

4-81 

+ 8l 



G 



POTASSIUM I 

FERRO- ' 

CYANIDE, I 

5 PER CENT I 



TANNIN, 
1 PER CENT 



-|-8t 

+ et 

+ 



- ' + 



I 



S 



+ 8l 
+ 8l 

-f 



+ 8t 



G 



4- 

-|-8t 



+ ? 



? - 



I — I _ 



"S" refers to bone aerum; "G" to glutia. Order of mixing solutions: acetic-acetate 
znizttire 1 cc., water 1 oe., reagent 1 drop and serum (0.5 percent) or glutin (0.5 percent) 1 drop. 



are made available for the precipitation reaction with the alka- 
loidal reagents. It is thus possible to ascertain if precipitation is 
concerned (1) with neutral particles of protein, (2) with positive 
protein ions, or (3) whether it depends upon the liberation of the 
free acid of the reagents. With this in view, experiments with 
horse serum and glutin were made according to the method de- 
scribed above, the details of which are appended to the results pre- 
sented in Table II. 

From this it is to be seen that the precipitation of serum by the 
following reagents, iodine potassium iodide, potassium mercuric 
iodide, potassium ferrocyanide, sodium tungstate, and phos- 



1 6 Precipitation of Serum-Albumin and Glutin 

phomolybdate, takes place above the isoelectric point, and in the 
direction of higher concentrations of acetic acid, and that it in- 
creases as the acidity increases. Precipitation by the different 
reagents begins practically with the same mixture of acid and 
acetate, and the small variations which occur are probably within 
experimental error. Inasmuch as the dissociated acid is not 
bound to the protein ion under these conditions, it would seem to 
indicate that a certain amount of free and excess of acid is neces- 
sary for the formation of the insoluble protein compounds. 

On the other hand, with tannin, the maximum of precipitation 
occurs at about the isoelectric point, and then diminishes on either 
side of it; that is, with either increased or decreased concen- 
trations of free acid. This would indicate that the mechanism 
of precipitation by tannin is different from that of the other alka- 
loidal reagents used. 

With glutin, precipitation takes practically the same course as 
with serum. No precipitation was observed with iodine and po- 
tassium iodide and the ferrocyanide. Tannin exhibits practically 
the same differences from the other reagents as with serum. 

2. Precipitation of serum and glutin by alcohols in mixtures of acetic 

add and acetaie. 

From the preceding section it appears that the mechanism of 
the precipitation of proteins by tannin is different from that with 
the other reagents. It is possible that the mechanism is similar 
to that of certain alcohols; for ethyl alcohol also produces a max- 
imum of precipitation at the isoelectric point. This analogy was 
tested out by observing the precipitation of serum and glutin by 
different alcohols with respect to the isoelectric point. The data 
are presented in Table III. 

It is seen that the maximum of precipitation of serum by the 
different alcohols used occurs at about the isoelectric point; i.e., 
with mixtures of one to two (i) and one to four (J) parts of 
acetic acid and acetate, respectively. With the exception of 
propyl alcohol, the precipitates diminish in intensity on either side 
of this point. There appeared to be no difference in the pre- 
cipitates produced by propyl alcohol. It is also to be noted that 



Paul J. Hanzlik 



17 



TABLE III. 
Precipitation of serum by alcohols in the acetic-acetate mixtures. 



1^0 ACBTIC ACID 



V/10 SODIUM ACBTATE 



PHENOL, 
5 PER CENT 



RESOBCIN, HTDROQUINOmB, 
5 PERCENT 5 PER CENT 



t 

I- 

1-s 



PBOPTL 
ALCOHOL 



\ 

1 

I- 

1 

J 

1 

J 



I 

IT 



Serum and water (equal 
parts) 



+ 

+ 8t 
-h 8t 

4- less st 



+ ? 

+ 

-l-st 

+ St 

+ less st 



I 



+ still less I + least st 
st 



-h st 
-h st 
+ less st 
-h still less 

st 
+ tr 



4-tr 
-hst 

+ 8t 

-hst 
4- St 



+ least st 4- least st 1 + tr 



Onkrof mixing solutions: hone serum (0^ per cent) 1 oc., acetic-acetate mixture 1 cc, and 
sleoboircicent 1 ce. 

serum alone without the acid-acetate mixture gave slight precip- 
itates with all alcohols except propyl alcohol. The data strongly 
indicate that the mechanism of precipitation by tannin more 
closely resembles the alcohols than the other alkaloidal reagents. 

The results obtained with glutin (Table IV) lead to practically 
the same conclusions. 

TABLE IV. 
f^ecipitation of glutin by alcohols in mixtures of acetic acid and acetate. 



X/IO ACETIC AQD 



V/10 SODIUM ACETATE 



PHENOL, 
5 PER CENT 



RESORaN, 
5 PER CENT 



f. 
f. 
}•• 

I 
J- 

1 
f- 



4- 



-f- less st 
+ st 
4- st 
4- V. si 



HTDROQUINONE, 
5 PER CENT 






PROPYL 
ALCOHOL 



+ 

4- 
4- 
4- 



Order of mizincsdutions: glutin (0.5 per cent) 0.5 cc.. acetio-acetat« mixture 0.5 00., water 
1 oCh sod alcohol 2 oe. 



nra JomurAL or aioLOoicAL cmnfisTRT, tol. xx, no. 1 



1 8 Precipitation of Serum- Albumin and Glutin 

3. Precipitation of serum and glutin in acids. 

Thus far it is seen that the precipitation of serum and glutin 
by the various reagents occurs when acid protein is present; that 
is, it takes place in the presence of a concentration of hydrogen 
ions higher than the isoelectric point, and, apparently, is aug- 
mented when these are in excess. However, it still remains to be 
shown whether precipitation depends upon the liberation of the 
acid of the reagent, or the formation of protein salts with added 
acid and the reaction of these with the reagent. 

According to Pauli, the dissociated hydrogen ions of such acids 
as acetic or hydrochloric are bound by protein with the formation 
of dissociated protein ions with positive electric charges. The 
number of these charges increases with the addition of more acid 
until all of the protein is saturated. This occurs even in the lowest 
concentrations of acid. Such bound hydrogen ions do not become 
available for other reactions unless the quantity of protein remains 
constant and an excess of acid is added. If, then, precipitation 
in a mixture of protein and acid by a reagent occurs only during 
the phase of excess acid and not when protein ions only are pres- 
ent, the conclusion must be drawn that precipitation depends upon 
the presence of free acid and not upon dissociated protein ions 
alone; that is, the precipitate is a combination of the free acid of 
the salt with the protein or acid protein. This, indeed, is the case 
with iodine potassium iodide, mercuric iodide, ferrocyanide, tung- 
state, and phosphomolybdate, and to a certain extent with tannin. 

The experiments were performed by adding small and constant 
quantities of serum and glutin to different concentrations of acetic 
aiid hydrochloric acids and noting where precipitation occurred 
by the further addition of constant quantities of different reagents. 
The data obtained, as well as the details in which the experiments 
were performed, are presented in Table V. 

It is seen that in the lowest concentrations of both acids, where 
dissociated protein ions occur, no precipitation of serum with any 
of the reagents took place. Precipitation occurred only with the 
higher concentrations; that is, in the presence of an excess of acid. 
The beginning of precipitation with the different reagents varies, 
and small but negligible variations are to be noted in the same 
reagent with different acids. With tannin, a maximum of pre- 



Paul J. Hanzlik 



TABLE V. 

Prtcipilatitm of aerum and glulin by alkaloidal reanenl* in dilute aoiih. 



llOUniH «"*™™ ' SODIUM ' BODItm fclTHMinS. 



ilH ,- 
T.1,1H - 
Wlfr* 1- 

<1H l+rt 


- 

-i-'l 

+ 


+ 
+ 
+ « 

4 


+ 

+ 

+ 

+ 


+ 
+ 

+ 
+ 
+ 
+ 


2 


- 1- ,- i- ;- 


+ r 

+ 

+ T 
+ 1 










+ 
+ 


- - - '+d +IOI.H 

+ ■] - 1 - ; +ri + 
+.1 - +1jl+l«lll + 

+ it ' + 1 + 1 + n +]mM. 









Otirtte. 










;iuH - 

W-lfr* [ - 

1.1.1*. - 

I.S 1(H 1 - 

tin 

1I.1M - 

I11.IIH - 
IIKH 1- 


: 


+ 

+ 
+ 
+ 


+ - 
+ ' - 
+ ' - 

+ +•! 
+ '+■! 


+ •1 

+ .1 


- 


- 


■s 

1 1 1 1 1 1 1 1 + 

■5 


+ 
+ •1 ' + 

+mor. + 
+ «illi + 

+ «lill' + 

i + 



oe., iJutia or horn • 



dpitation in both acids occurred at the concentration of 2. 5. 10-*. 
In the higher concentrations the precipitates dissolved. This was 
also true of iodine and potassium iodide, which at no time gave 
ibeolutely distinct precipitates. This phenomenon perhaps de- 
pends upon the formation of soluble acid salts and requires further 
investigation. 
With glutin, practically the same results were obtained as with 



20 Precipitation of Serum-Albumin and Glutin 

It is to be concluded that the precipitation of serum and glutin 
in hydrochloric and acetic acids by iodine potassium iodide,' mer- 
curic iodide, ferrocyanide, tungstate, and phosphomolybdate, takes 
place only in the presence of an excess of free acid and depends 
upon the formation of insoluble compounds with the free acid 
of the salt. 



4 



The effect of salts on the precipitation of serum and glutin by 

aUcaloidal reagents. 



The addition of salt to an acid might conceivably alter the pre- 
cipitation reaction by alkaloidal reagents and alcohols. This was 
studied in the following manner: To constant and de£biite con- 
centrations of glutin and serum in the same volumes of acetic and 
hydrochloric acids of different concentrations were added con- 
stant volumes of potassium chloride and potassium sulphocyanide, 
and finally to this mixture constant and definite quantities of the 
reagents. Each series of experiment^ with the different salts was 
compared with a series without salt; that is, with distilled water 
as a blank. The results from all of the experiments are to be 
found in Tables VI, VII, and VIII. 

From these data it is seen that the two salts, chloride and sul- 
phocyanide, exerted practically no influence upon the formation 
of the precipitate by the various reagents used. Practically no 
precipitation occurred in any case until the concentration of 4 . 10""* 
was reached. The small variations which occurred fall within 
the experimental error. This was true of both serum and glutin. 

With tannin, precipitation in the presence of the chloride and 
sulphocyanide occurred in the lowest concentrations of the hydro- 
chloric acid and before 4.10^. Thus a small difference from the 
other reagents is indicated. 

With alcohols, the addition of salts also had practically no in- 
fluence upon the precipitation. The sUght variations exhibited 
by the sulphocyanide are too small to permit the drawing of any 
conclusions. The same can be said of glutin in acetic acid-ace- 
tate mixtures (Table IX). 



»- + + + + + + 



I " i 



h++ + +++ 






«'8 I « I S 

++++ + ++ 



I 

9 






f++ + +++ 



I Pi- 

f is H^ 



III 



■<-+ + + + + 



I + + + + + +4 



Mt3 



spimi^ 



11 i 



25ii J 



+ + + + + + + + + + + I 



+ + + + + + + + + + + I 












I 



++++++++++++ 



++++++ 



I 

:| 

•3 
••• 

SI 

•c 

I. 



i 



M 

e 



z 



<2 



i 



h 



u 



8e 



M 

H 



2 

S 

M 






n 



H oG 

*• ■ M 

s3-» 









gi 






•0 



75 ■ 



H ^ 
X a, 

D - 









I *^ 



B 
I, 

M »4 



^ I 



M 
Q 

H 

is 

Q 
O 



I I I 






M I 






I I I 



I I I 



+ + + + + + + + + + + I 



+ + + + + + + + + + 4- I 



I + 



I + 



I + 



I + 



I + 



fl^ ! 



I + 



+ + 



en ■ 
+ + 



+ + + 



+ + + + + 



++++++ 



++++++ 



I I I I I I 



I I I I I I 



I I I I I I 



I 









U."- 






C4 



• . • O • o 



Q 

§ I 






I + 



m 
I + 



I + 



e 

a 
, B 



e 

s 



I I 



I I 



+ + 



• Bi » 
+ + + 



a 

9 

-a 

8 

e 
.S 
t 

1 

z 

.g 

1 

o 



X ■ 
I + + 



+ + + + + 



+ + + + + 



I I I I I 



I I I I I 



I I I I I 






• • 

I 

9 

8 






22 



• • 



Paul J. Hanzlik 



TABLE VIII. 

Efftet of tallt on Ike precipitation of terttm and glatin by alcohol*. 





SnHnirr 


.■^S?;;; 


B Ptn ram 


PK)PTi.<un>aL 


'"■^^•^ 


3 


i 


z 


g 


g|i 


s 

- 

+1 


§ 


g 


8 


B 


i 


l.M.IO-1 
O.I.IIH 
B.l.lt-> 

n.t.iB-' 

I.U.lO-' 
0« l(H 
1 U.IO-< 

I.U.llh* 

l.U.MH 


+ 




+ 
4- 


+•1 


+•1 


+ 
+ 


+ 


+ 
+ 


4.1 


4«1 


+•1 


ScwacAca, 


























(H.HM 
IK.IO- 

I.M10-i 

0.«.lD-> 
1»1CM 


+ 




+ T 

+ M 

+ 


+ 




4- at 
+ <1 


i 




+ 


+ 




+ 



SCI « KSCN) (oluUo 



d 0J> «., tfinio (0.1 p«r a 



Efftel vf ntlphoeyanide on the precipitation of glulin by atcohoh in acetic- 
acetate mixtures. 



>/10 Mxnc uut 


.^SS t .- 


'^W fe2^<^" 


XKOHOL 


>/10K>M<niAcnAi. 


BiO 


kbcn; H.0 


KSCN HiO 


KSCN 


BiO KBCN 


( 


+ al 
+ at 

+ 8t 

+ sl 


+ at - 
+ at ! + at 
+ al + St 


1 1 1 1 I++ 


+ 


+ 






+ Bt 
+ 


+ ? 

+ ? 
+ f 
+ ? 






A 


_ ^_ 1_ 



24 Precipitation of Serum-Albumin and Glutin 

B . Effed of concentration of protein on the precipitation. 

ThiH was tested out by usii^ two different concentrations of 
aerum, 0.017 per cent and 1 per cent (as end concentrations of 
protein) in constant volumes of hydrochloric acid of different 
atrengths and precipitated by the further addition of one drop 
of the reagent. The data have been placed in Table X. 

As the results with the different concentrations of scrum are 
practically identical, it is to be concluded that wide differences 
in protein content have no marked influence on the precipitation. 

TABLE X. 

Effect of concentration of eeram-aibvrain on the precipitation by 
alkaloidal reagents. 



B.IO-* 
lO.IO-* 

.B.IIH 
.1.10-' 

10. lO-" 
J.KH 
».10-< 



inn 


■ODIITK 


roTunuu 




























Swu 









.. wid reMe°^ ' drop. The : 



CDD«ntratioiu of ^buouD. 



The mechanism of the precipitation of dialyzed horse serum and 
glutin by tannin is different from that of certain precipitants com- 
monly known as "alkaloidal reagents." With these a certain 
amount of free acid (hydrogen ion concentration) is necessary 
for the formation of the complex msoluble compounds. 

Tannin behaves like certain alcohols, e.g., r^sorcin, phenol, 
hydroquinone, and propyl alcohol, since the maximum of precip- 
itation in both cases corresponds to the isoelectric point in serum- 
albumin and glutin. 

Precipitation of scrum is uninfluenced by wide differences of 
concentration and the addition of such neutral salts as chloride 
and sulphocyanide. 



SIMPLER NUCLEOTIDES FROM TEAST NUCLEIC ACID. 

Br WALTER JONES and A. E. RICHARDS. 

(From the Laboratory of Physiological Chemistry, Johns Hopkins 

University, Baltimore.) 

(Received for publication, November 9, 1914.) 

It has been showo by Levene and Jacobs* that when yeast 
nucleic acid is boiled with dilute mineral acids, the tetra-nucleotide 
is decomposed into its four component mono-nucleotides, as is 
shown in the following diagram. 

HO 

0=P-0 CHgOs-CsHiNsO 1 

/ 
H 

HO "\ 

0=P-OC,HgO, C4H4N,0 2 

/ 
H O 

HO' \ 

= P-0 C6HgO,CiJl4N6 3 

/ 
H 

= P - CsHgOs C4H,N202 4 

/ 
HO 

Yeast nucleic acid 

But at the same time the two purine mono-nucleotides (1 and 3) 
are further decomposed by the boiling mineral acid into purine 
bases, phosphoric acid, and pentose, while the two pyrimidine 
mono-nucleotides (2 and 4) are unaltered and may be obtained as 

• P. A. Levene and W. A. Jacobs: Ber. d. deulsch. chetn. Gesellsch., xliv, 
p. 1027, 1911. 

2S 



HO 



26 Nucleotides from Yeast Nucleic Acid 

end products of the hydrolysis. This is the first recorded instance 
of the formation of simpler nucleotides from a tetra-nucleotide, 
and the results furnish a method by which one can easily decide 
whether in a given instance he is dealing with a purine nucleotide, 
a p3aimidine nucleotide, or a mixed nucleotide containing both a 
purine and a pyrimidine group; for in such a case it is not neces- 
sary to make a tedious purine-pyrimidine separation, but is suffi- 
cient to estimate the phosphoric acid liberated by hydrolysis with 
dilute mineral acid. 

In a former contribution' we stated that the first stage in the 
enzymatic hydrolysis of yeast nucleic acid consists of a decomposi- 
tion of the tetra-nucleotide into two di-nucleotides, as is shown in 
the following diagram. 

HO 
= P- • CjHgO, • C,H4N,0 



/ 



/ 

H O 



Guanine -cytosine 
\ di-nucleotide 

O = P - • Cai,0, • C«H4N,0 



HO \ 

= P- CHjO, • C»H4Nj 

/ 

O Adenine-uracil 

\ di-nucleotide 

= P - O • CjHgO, C4H,N,0, 

/ 
HO 

Yeast nucleic acid 

A brief statement of the crucial properties of the di-nucleotides 
was made; ws., that by acid hydrolysis the one yields essentially 
guanine and cytosine, while the other yields adenine and uracil; 
but full discussion of the substances was postponed. We showed 
that when the enzymatic hydrolysis of yeast nucleic acid proceeds 
farther, the guanine-cytosine di-nucleotide is decomposed into it3 
component mono-nucleotides. One of these was found among 
the end products of the action of yeast on yeast nucleic acid and 

* W. Jones and A. E. Richards: this JoutnaXy xvii, p. 71, 1914. 



W. Jones and A. £. Richards 37 

was shoTrn to be identical wittt the guanylic acid which is widely 
present in animal glands, but especially in the pancreas. 

HO 

O = P - O • C»HeO, • C»H4K0 GuanyUc acid 

/ 
H O 

"HO \ 

O = P- O C^,0, • CANjO 

Guan ine-cytosine. di-nucleot ide 

Since the appearance of our former paper the field has been con- 
siderably broadened by an important communication of Thann- 
hauser,' who found that yeast nucleic acid is stripped of one of its 
nucleotide groups by the digestive action of duodenal juice, and 
there is formed a tri-nucleotide which contains groups of guanine, 
adenine, and cytosine, but not uracil. Thus six simpler nucleo- 
tides have been prepared from yeast nucleic acid, — a tri-nucleotide, 
two di-nucleotides, and three mono-nucleotides. 

It is our purpose to discuss the two di-nucleotides and guanylic 
acid in this paper. 

The extent to which enzymatic hydrolysis of yeast nucleic acid 
will proceed depends upon a number of factors. Various animal 
and plant tissues differ greatly from one another in the ferments 
which they contain, so that, under precisely the same conditions, 
extracts of different tissues will produce different end products by 
their action on the nucleic acid. But the same tissue extract may 
produce quite different end products by a change of the conditions 
under which its enzymatic activity is exerted. One of the princi- 
pal factors in this connection is the relative amount of nucleic acid 
that is initially brought into contact with the ferment solution. 
Thus an aqueous extract of pig's pancreas by its action on nucleic 
add may be made to yield either guanylic acid, guanosine, guanine, 
or xanthine, by diminishing successively the amount of nucleic 
acid employed with a constant amount of the pancreatic extract. 
Finally, the enzymatic decomposition of nucleic acid is brought 

* S. J. Thannhauaer: Ztachr. /. phyaiol. Chem,, xci, p. 329, 1914. 



28 Nucleotides from Yeast Nucleic Acid 

about by the successive action of a large number of independent 
agents or enzymes which differ markedly from one another in their 
stability at the body temperature, so that, by submitting a tissue 
extract to digestion at 40° certain of its ferments may be destroyed, 
while others remain in considerable activity, and therefore the end 
products of the action of such a predigested extract upon nucleic 
acid will be different from those of a fresh extract of the same tissue. 
Advantage has been taken of this last circumstance in the prepa- 
ration of the two di-nucleotides from yeast nucleic acid. If a fresh 
aqueous extract of pig's pancreas be allowed to act on yeast nucleic 
acid under the conditions stated below, the end products (as far 
as concerns the guanine group of the nucleic acid) will be guanylic 
acid and guanosine. The two substances will combine with one 
another to form an insoluble compound whose separation from the 
other products is easily accomplished by the method described in 
connection with yeast.* This decomposition of nucleic acid evi- 
dently proceeds with the formation in turn of: 

1. Guanine-cytosine di-nucleotide. 

2. Guanylic acid. 

3. Guanosine. 

For if the pancreatic extract be digested at 40° before adding 
the yeast nucleic acid, the ferments that decompose the di-nucleo- 
tides are destroyed, while the ferment that produces them from 
nucleic acid remains active, so that the two di-nucleotides com- 
pose the principal end product of the action of this extract on yeast 
nucleic acid. 

These two substances have general properties somewhat similar 
to those of yeast nucleic acid, and like yeast nucleic acid they 
easily undergo decomposition when submitted to processes in- 
tended for their purification. It should not therefore be expected 
that they have been prepared in perfectly pure condition; but if 
the very fair approach of their chemical properties to theoretical 
requirements be taken into consideration with their origin from 
yeast nucleic acid, the structure of the compounds follows with 
considerable certainty. 

* Jones and Richards: loc. cit. 



W. Jones and A. E. Richards 29 

Preparation of the two di-nucleoiides from yeast nucleic acid. 

A mixture of 2 kilos of carefully trimmed and ground pig's 
pancreas, 2 liters of water, and 30 cc. of chloroform was allowed to 
digest for twelve hours at the room temperature in a tightly closed 
vessel with frequent and violent agitation. After the tissue had 
by this means become thoroughly penetrated with chloroform, 
the mixture was placed in a thermostat and allowed to digest at 
40** for two weeks, when it was cooled and filtered. The clear, 
pale yellow filtrate was then treated with yeast nucleic acid (1.5 
grams for each 100 cc. of fluid) and again digested at 40° for twelve 
hours, when the complete disappearance of the nucleic acid was 
apparent from the failure of the clear fluid to produce a precipitate 
or even a cloud with sulphuric acid. The product was heated 
to boiling, filtered from a small coagulum, and treated at the boil- 
ing point with a 25 per cent solution of lead acetate as long as the 
reagent produced a precipitate in the hot fluid; but care was 
taken to avoid a great excess of lead acetate, which would precipi- 
tate the di-nucleotides. It is not at all difficult to make this 
precipitation accurately, because there is a wide margin between 
the point where the precipitation of lead phosphate, coloring mat- 
ter, etc., ends, and the point where the precipitation of the di- 
nucleotides begins. A number of objectionable substances are 
thus removed, leaving a clear, pale yellow fluid which can easily 
be filtered from the bulky lead precipitate, and which forms a 
heavy, granular precipitate when treated at the boiling point with 
an excess of lead acetate. This lead precipitate increases in 
amount as the fluid cools, and settles so rapidly that the super- 
natant liquid may be removed by decantation. It consists of a 
mixture of the lead salts of the two di-nucleotides. By suspending 
the lead salts in warm water and treating with hydrogen sulphide, 
a solution of the free di-nucleotides is obtained. 

The method of separating the two substances from one another 
is based upon the peculiar conduct of their potassium salts when 
treated in aqueous solution with an excess of alcohol, the one sub- 
stance being inamediately precipitated, while the other remains 
emulsified. Accordingly, the aqueous solution of the di-nucleo- 
tides was filtered from lead sulphide, concentrated under dimin- 
ished pressure at 45°, and treated warm with a small amount of 



30 Nucleotides from Yeast Nucleic Acid 

hot concentrated potassium acetate. The potassium salts are 
slowly formed and the solution becomes gelatinous as it cools, 
but before gelatinization has proceeded far, the warm aqueous 
solution is stirred into an excess of alcohol. The guanine-cytosine 
di-nucleotide is immediately thrown out as a perfectly white floc- 
culent precipitate, which falls rapidly, while the adenine-ut-acil di- 
nucleotide remains suspended for a long time, but on standing over 
night becomes deposited as a resin leaving a perfectly transparent 
fluid. This resin which entraps all the coloring matter of the solu- 
tion may be hardened by treatment with absolute alcohol and 
finally ground to a pale yellow, dry powder. 

It would be surprising if this process should effect a sharp sepa- 
ration of the two substances and one might expect the properties 
of each compound to be expressed in the other. Nevertheless, the 
separation is sufficiently accurate to establish the most important 
property of both compounds; t. 6., that each 3rields principally but 
one purine derivative and but one pyrimidine derivative. 

The guanine-cytosine di-nudeotide. 

The crude substance obtained as described and dried with ab- 
solute alcohol was made into a hot concentrated solution and 
treated with an equal volume of glacial acetic acid which produces 
a bulky gelatinous precipitate. This was filtered ofif and the fil- 
trate was poured into a large volume of glacial acetic acid. The 
precipitated di-nucleotide was filtered off as closely as possible with 
a pump, dissolved in a little hot water, and after the addition of a 
small amount of potassium acetate was poured into a large excess 
of alcohol. The precipitated di-nucleotide was rapidly and 
roughly washed free from the cloudy solution by decantation with 
absolute alcohol, dried in a desiccator with sulphuric acid, and 
finally heated to a constant weight at 105°. The substance was 
thus obtained as a perfectly white non-hygroscopic powder, easily 
soluble in water and laevorotatory to polarized light. 

Two grams dissolved in 50 cc. of water gave a reading of — 1°36' 
in a 2 dm. tube (a)D= -20^ 

Yeast nucleic acid, on the contrary, is strongly dextrorotatory. 
Analysis of three different preparations of the di-nucleotide gave 
the following percentages: 



W. Jones and A. £. Richards 



31 







FOUND 




REQUIRED FOR 




I 


U 


lU 


CttHMPtNlOlOCt 


Potassium 

Nitrogen 

Phosphorus 


9.9 
15.2 

7.8 


10.2 
15.6 

7.8 


10.4 
15.5 

7.7 


10.8 

15.0 
8.3 



Several weighed portions of the di-nucleotide, 1 gram each, were 
heated with 10 cc. of 5 per cent sulphuric acid for an hour in a 
small flask provided with a simple condensing tube and immersed 
in boiling water. The product while still hot was treated with 
an excess of ammonia, and, after cooling, the coarsely granular 
guanine was filtered off, dried, and weighed. The ammoniacal 
filtrates were then treated with magnesia mixture for the pre- 
cipitation of phosphoric acid, and the filtrates from magnesium 
ammonium phosphate were tested for purine bases with an am- 
moniacal solution of silver nitrate. In case a small precipitate 
fonlied it was decomposed and tested for adenine. 

The specimens of magnesium ammonium phosphate were dis- 
solved in nitric acid and precipitated in turn as ammonium phos- 
phomolybdate and ammonium magnesium phosphate. The latter 
were weighed and the phosphorus was calculated. It will be seen 
from the results tabulated below that only about half the total 
phosphoric acid of the compound is liberated thus by mild acid 
hydrolysis, which shows that the substance is a mixed di-nucleo- 
tide containing both a purine and a pyrimidine group. 

Each specimen of guanine, after weighing, was converted into 
the chloride which consisted uniformly of the feathery, macro- 
scopic needles characteristic of guanine chloride. . Some of the 
specimens were analyzed. 

The results of five such experiments made with different prepara- 
tions are exj^ressed as percentages in the following table. 



32 



Nucleotides from Yeast Nucleic Acid 



OUAKIXE 



PHOSPHORUS 



I. 

II. 

III. 

IV. 

V. 



1 

1 

JTheoretical 

1 
1 


Found 
21.7 


Theoretical 

for yeast 

nucleic 

acid 


▲DKNime 


- — 

Total 


Half 

1 

4.2 


20.3 


10.5 





8.3 




20.9 
21.3 
22.0 






Trace 








"1 


21.1 




Trace 







Found 

3.2 
3.4 
3.3 
3.6 
4.0 



Thirty grams of the di-nucleotide were heated for three hours 
with 25 per cent sulphuric acid in an autoclave at 150°, and the 
product was examined for pyrimidine derivatives by the well known 
silver method. Four and a half grams of characteristic cytosine 
picrate were obtained, from which the free base was prepared, 
dried at 115°, and analyzed (N = 37.42 and 37.61, instead of 
37.85). Uracil could not with certainty be identified among .the 
products. A trace of a pyrimidine derivative was obtained out 
of the filtrate from cytosine picrate, but it could not be determined 
whether it was cytosine or uracil. Its amount was too small to 
be of any significance. 



The adenine-uracil di-nucleotide. 

The crude pale yellow powder was purified by means of glacial 
acetic acid as described above; but the substance could not be 
obtained in as pure a condition as the di-nucleotide discussed in 
the previous section. Its decomposition products showed it to 
be contaminated with considerable traces of this substance, as 
might have been expected from the crude, but only available 
method of separating the two substances from one another. But 
this di-nucleotide fulfils its two most necessary requirements. It 
Uberates approximately one-half of its phosphoric acid by mild 
acid hydrolysis, and jdelds nearly twice as much uracil as can be 
obtained from a corresponding amount of yeast nucleic acid. 
From 35 grams of the substance by hydrolysis with 25 percent sul- 
phuric acid in the autoclave at 150°, there were obtained 4 .08 grams 
of pure recrystallized uracil in characteristic needle clusters. 
(N = 25.01, 25.11, 25.07, 25.03, instead of 25.00.) 



W. Jones and A. E. Richards 



33 



The di-nucleotide is slightly laevorotatory to polarized Ught. 
Two grains of substance in 50 cc. of water gave a reading of — 36' in 
a 2 dm. tube (a)D= -7.5°. 

A number of specimens of the di-nucleotide, 1 gram each, were 
heated on the water bath for an hour with 5 per cent sulphuric 
add. The hot solujbion was made strongly alkaline with ammonia, 
and the small amount of guanine which was deposited when the 
solution cooled was filtered off. The filtrate was then examined for 
adenine and free phosphoric acid. The results of five such experi- 
mcDts are expressed as percentages in the following table. 



OUAinNB 



ADBNINE 



PBOSPROROa 



Theo- 
retical 



I. 

II. 

III. 

IV. 
V 







Found 



Theo- 
retical 



Found 



1.2 
1.0 
0.8 
0.2 
1.1 



20.4 



Theo- 
retical 

for 

yeast 

nucleic 

acid 



Total 



Half 



Found 



13.7 


9.6 


9 37 


4.68 


16.2 






12.8 


' 




14.6 


1 




14.7 









4.01 
4.12 
4.13 
3.96 
4.21 



These results show that the material was somewhat contami- 
nated with the guanine-cytosine di-nucleotide, or with some unde- 
composed nucleic acid. But the low values for adenine are of 
little significance; for no better results for adenine can be obtained 
with the purest preparations of yeast nucleic acid.* It is probable 
that the analytical piocedure for adenine is not as exact as is com- 
mcmly supposed. 

Guanylic acid. 

In our former contribution we showed how yeast nucleic acid is 
decomposed by the ferments of yeast with the production of a com- 
pound of guanyUc acid and guanosine whose difficult solubiUty 
makes it easily accessible. From this substance (which is formed 
abo by the action of fresh pancreas extract on yeast nucleic acid) 
guanylic acid may best be prepared in pure condition by the fol- 
lowing procedure. The substance is dissolved in a comparatively 

*See results of P. A. Levene: Biochem. Ztschr.^ xvii, p. 120, 1909. 



TBSJOVBirAI.Or BTOLOOICALCHXIOSTIIT, VOL. XX, KG. 1 



34 



Nucleotides from Yeast Nucleic Acid 



large amount of hot water and treated with a solution of lead ace- 
tate. To prevent any guanosine from being carried down with 
the heavy granular lead guanylate, the precipitation is made slowly 
and at the boiling point. After thorough washing with boiling 
water the lead compound is suspended in cold water and decom- 
posed with hydrogen sulphide. The excess of the reagent is driven 
out of the cold solution by a current of air and the fluid is evapo- 
rated to a small volume at 50° under diminished pressure. Upon 
cooling in ice water, perfectly white guanylic acid is deposited. 
This sometimes contains a trace of guanosine which may be 
removed by dissolving the guanylic acid in cold water (which 
leaves the guanosine undissolved) and precipitating the material 
with alcohol. A very pure product may also be obtained by the 
addition of alcohol to a mother-liquor from which a little guanylic 
acid has been deposited. Analysis of seven different preparations 
precipitated with alcohol and dried at 105° gave the following 
results: 

I. 0.2783 required 14.58 cc. of standard acid (1 cc. =0.003695 N). 

II. 0.1614 required 8.24 cc. 

III. 0.2118 required 10.88 cc. 

IV. 0.5351 gave 0.3498 MgNH4P04.6H20 
V. 0.5415 gave 0.3571 MgNH4P04.6H20 

VI. 0.5238 gave 0.3412 MgNH4P04.6H20 
VII. 0.4644 gave 0.3000 MgNH4P04.6H20 





KOUND 








THEORETICAL 




1 ! 11 


lU 1 IV 

18.98; 

8.36 


V 


VI 

8.27 


Vll 
8.51 


FOR 

CioHMNa>Ot 


N 


19.35 18.86 

1 
1 


8.37 


19.28 


P 


8.55 






i 





The formation of guanylic acid from yeast nucleic acid suggests 
the explanation of an interesting and curious matter. The tetra- 
nucleotide which is a constituent of animal cell nuclei, z.e., the so 
called animal nucleic acid, contains a hexose group. This can 
scarcely be doubted, for by severe acid hydrolysis the substance 
yields both formic acid and lacvulinic acid, and by oxidation with 
nitric acid it forms one of the saccharic acids. On the other hand, 
the tetra-nucleotide of plant cell nuclei or plant nucleic acid pos- 
sesses a pentose group, and pentose is formed when the substance 



W. Jones and A. E. Richards 35 

is hydrolyzed. Guanylic acid is a mono-nucleotide present in 
animal glands and might be expected to resemble in its chemical 
structure the tetra-nucleotide of animal tissues. But this is not 
the case. Guanylic acid contains a pentose group, not a hexose 
group, and by acid hydrolysis it forms not only a pentose but the 
particular pentose which is formed similarly from plant nucleic acid 
(d-ribose;.* The idea is at once suggested that the guanylic acid of 
animal tissues is formed from the nucleic acid of plant food, and 
this idea is strongly supported by the actual production of guanyUc 
acid from yeast nucleic acid by the action of fermerts. 

In order to throw additional light upon this question we under- 
took the quantitative preparation of guanylic acid from the pan- 
creas of different animal species, the ox and the pig. The sub- 
stance cannot be obtained in very great amount from the pancreas 
of either species, and the methods in use for its isolation are far 
from what might be desired; so that a sUght difference observed 
would be without significance. But we obtained about three and 
one-half times as much guanyUc acid from ox pancreas as from an 
equal weight of pig pancreas. 

All things considered, we are incUned to believe that the guanyUc 
acid of animal tissues has no direct relation to animal nucleic acid, 
but owes its origin entirely to the nucleic acid of plant food. 

• I^vene and Jacobs: loc. cit. 



STUDIES ON FASTING FLOUNDERS.' 

By SERGIUS MORGULIS. 
{From the United Slates Fisheries Biological Station^ Woods HoUj Mass.) 

(Received for publication, November 16, 1914.) 

In a series of investigations on the nutrition of aquatic animals 
Putter* propounded the theory that the sea contains considerable 
quantities of organic matter in solution serving as nutrients for 
these organisms. The energy required by the aquatic animals 
is, according to this view, derived from two sources: from the food 
which is ingested in the ordinary manner, and from the dissolved 
organic matter which, in the case of fishes, is supposed to be ab- 
sorbed through the gills. 

Simple as this theory may seem in its fundamental proposition, 
the arguments evolved by Putter in the development of his thesis 
rest upon an intricate network of assumptions and offhand cal- 
culations. That the gills might act as an organ of absorption is 
in accordance with their embryological history; but, as Moore- 
points out, we know of no substance in the gills which, by virtue 
of its strong affinity for the organic material in solution, would 
remove this material from very large volumes of water. But. 
whatever may be surmised regarding the gills. Putter's contention, 
as far as the experimental proof is concerned, has been shattered 
at every point. 

Putter's claim that sea water contains large quantities of car- 
bonaceous material in solution has been shown by Henze* to be 

^ Published by permission of the Commissioner of Fisheries. 

? A. Piitter: Die Ern&hrung der Fiscbe, Ztschr. f. allg. Physiol., ix, p. 
147, 1909. 

» B. Moore, E. S. Edie, E. Whitley, and \V. J. Dakin: The nutrition and 
metabolism of marine animals in relationship to (a) dissolved organic mat- 
ter and (b) particulate organic matter of sea water, Biochem. Jour., vi, pp. 
255-296, 1912. 

* M. Henze: Bemerkungen zu den Anschauungen Ptktters Ubcr den Ge- 
halt desMeeres angeldsten organischen Kohlenstoffverbindungen und deren 
Bedeutung fCir den Stoffhaushalt des Meeres, Arch. /. d. ges. Physiol., 
cxxiii, p. 487, 1908. 

37 



38 Studies on Fasting Flounders 

based on a faulty technique in working with the Messinger wet 
combustion method. The quantities which Henze's careful chem- 
ical analyses of the sea water showed were well within the limits 
of experimental error. More recently Moore^ showed that also 
in the application of the permanganate method for the estimation 
of organic material Putter committed serious analytical errors, 
having mistaken the reduction caused by an excess of chlorides 
for an indication of an abundance of organic matter. Besides, 
Moore worked with thoroughly filtered sea water, whereas Putter 
seems to have entirely overlooked the importance of this pre- 
liminary requisition. The results which Moore found agree ab- 
solutely with Henze's, both investigators having shown that the 
organic matter in the sea water is in practically negligible amounts. 

In upholding his thesis Putter resorted to another line of reason- 
ing, to prove his point by inference. He analyzed the classical 
case of tfie Rhine salmon, which, as Miescher has shown, sub- 
sists wholly on its own body reserves during the several months 
which it spends in fresh water. The energy which the salmon 
should expend, according to Putter's calculations, in swimming 
up-stream could not be derived from the oxidation of the material 
which the salmon loses in the course of the few months of fasting. 
Hence the conclusion follows that another source of energy must 
exist, which is the dissolved organic matter absorbed through the 
gills. This inference would claim some consideration if the energy 
requirement of the salmon had actually been determined; but, as 
a matter of fact. Putter estimates it from some experiments he 
performed with flounders; in other words, he bases it on another 
inference. Furthermore, as will be shown, the method employed 
in studying the energy output of flounders was such as to yield 
high values. 

By far the most valuable evidence which Putter claims in cor- 
roboration of his views has been obtained from a series of experi- 
ments in which the oxygen consumption of several kinds of fish 
has been determined during complete fasting. A study of the 

^ B. Moore, E. S. Edie, and E. Whitley: The nutrition and metabolism 
of marine animals; the rate of oxidation and output of carbon dioxide in 
marine animals in relation to the available supply of food in sea water, 
Report for 1913 on the Lancashire sea-fisheries, xxii, pp. 297-320, Liver- 
pool, 1913. 



Sergius Morgulis 39 

chemical composition of these fish at the close of the fast enabled 
him, by comparing it with the initial composition, to find out the 
loss in each constituent. From this it was a simple matter to cal- 
culate the amount of oxygen necessary to bum completely the 
lost material. The oxygen consumption of the fish has been found 
to be invariably greater than would be necessary for the combus- 
tion of this material, and this result again was considered by Piit- 
ter as positive proof that other substances, besides those lost by 
the animal, had partaken in the metabolism. Though this reason- 
ing seems logical, the argument fails to convince because the. total 
oxygen consumption of the fasting fish has not been actually meas- 
ured. Putter performed daily an experiment lasting thirty to sixty 
minutes; and from the results thus obtained he calculated how much 
oxygen the fish required for twenty-four hours. Any one familiar 
with respiration work cannot fail to appreciate that this procedure 
could only misrepresent the facts. When first put into the jar 
the fish on which the experiment is to be performed are generally 
restless for some time, and this is sujfficient to increase their re- 
spiratory exchange frequently by 50 per cent. In an experiment of 
long duration this temporary excitement of the fish has only a 
slight effect on the final results, but in experiments of only thirty 
minutes or one hour very high values can be found. Putter's con- 
tention that the oxygen consumption of his fish for the entire fast- 
ing period was more than that necessary to oxidize the body ma- 
terial is evidently based on an error which has been regularly com- 
mitted in the experiments and has been augmented through mul- 
tiplication by a large factor. Indeed, Lipschiitz* showed that the 
longer the daily experiments were made, the lower was the cal- 
culated amount of oxygen consumed in twenty-four hours. 

It seemed to me worth while, in order to settle this question 
definitely, to undertake a study of the oxygen consumption of 
fasting fish, avoiding as far as practicable all computations. With 
this in mind I performed experiments with small flounders whose 
oxygen consumption was measured continuously throughout the 
fasting period, except for the few minutes which were lost each 
time while changing the water of the jars. The two experiments 
reported in this paper lasted each 671 hours, and the oxygen in- 

• A. Lipschutz: tJber den Hungeretoffwechsel der Fische, Ztschr. f. allg. 
Physiol., xii, pp. llS-124, 1910. 



40 Studies on Fasting Flounders 

take was determined for almost 669 hours. In the cocirse of four 
weeks, therefore, the oxygen consumption for two hours only was 
not directly measured but calculated. 

At the end of the experiment both flounders were analyzed and 
the water, nitrogen, and fat contents determined. Their initial 
composition, at the beginning of the experiment, was figured out 
with the aid of the average percentage composition worked out 
for similar flounders. Owing to the smaUness of the sample the 
inorganic content of the material from the fasting flounders was 
not determined. 

METHODS.^ 

The oxygen consumption of the flounders has been determined by the 
difference in the oxygen content of the water at the beginning and end of 
an experiment. The sea water was aUowed to run into the jars, which 
had a capacity of about 4| liters, through a tube reaching down to the bot- 
tom. A sample of the water from the jar was siphoned into a small bottle 
of known volume. The stream through the siphon was maintained con- 
tinuously for about five minutes, so that the contents of the sample bottle 
were changed at least twenty times. The jar was closed hermetically by 
a vaselined plate. At the end of the experiment the plate was removed 
and a new sample of the water taken with the siphon which extended to 
about the middle of the jar. The water was drawn about two minutes, 
the contents of the sample bottle being thus changed about five times. 
The water was analyzed for oxygen according to Winkler's well known 
method. AH analyses have been made in duplicate, and the results have 
agreed closely. Since the exact volume of the sample and jar, and the pro- 
portion of oxygen dissolved in the sea water at the beginninj; and end are 
known, the oxygen consumption for a definite period is easily calculated. 
A special correction has been made on the basis of blank experiments. 

The analyses of the floimders were made in the ordinary way. The ani- 
mals were washed in distilled water and weighed as soon as all the super- 
ficial water was removed. They were desiccated in the open air, then 
ground up finely, and dried to constant weight. The material was ex- 
tracted with ether, the residue being weighed as fat. Part of the material 
after extraction was used for nitrogen determinations by the Kjeldahl 
method, and the rest was incinerated to find the ash content. 

The tables contain a complete record of the results of these 
experiments. It will be seen that the first flounder, D, weighing 
5.29 grams at the beginning of the experiment, i.e., twenty-four 

^ The methods employed have already been given in detail. S. Morgulis: 
Contributions to the gaseous metabolism of fishes, Proc, of the Biochem. 
Assoc. f Biochem. Bull., 1915 (in press). 



Sergius Morgulis 



41 



hours after the last feeding, consumed 0.192 cc. of oxygen per gram 
per hour, or 0.026cc. per square centimeter per hour. Tbifi animal 
was ver>' restless at first, but afterwards, and especially towards the 
end of the experiment, it kept very quiet. The weight was ascer- 
tained every four days. The total loss was over one-third of the 
original weight (34.4 per cent), when I was obliged to sacrifice the 
animal. In that time (668 hours, 40 minutes) it used up 485.56 

TABLE I. 



libU' :».o 






n-M, 



nbuy 

Ufa 10' 
ltb»' 



tt-v 

I7-M, 



»hiy 21.0 I 1T.S7 0.170 



Uovnalxiut 



42 



Studies on Fasting Flounders 





— - 


TAl 


8T.F. 11. 

2 ! oxi 






— 


1 






'OEM CON- 


1 




EXPERIMENT 




1 








»3 SUMEO 




2 


H 

p 


gin- 
ning 

1 


End 
a.m. 


Duration 


S3 

§ Total 

IS ' 


Per Per 
hour hour 
per per 
gm. |Sq.cm. 


REMARKS 


July 


gm. 


8q.cm. 


1 

' p.m. 




Cm CC. 


ce. ce. 




13-14 


4.04 


34.7ei 12:55 
, a.m. 


10:05 

j 


21h 10' 


20.5 15.51 

[ 


0.1811 0.021 Lost feeding July 12. 


14-15 






\ 10:10 


9:05 


22h55' 


20.5 ' 16.54i 




Animal is very quiet. 


15-16 


9:10 


8:10 


23h 


20.5 


10.33 


Animal is very quiet. 


16-17 






8:15 


9:55 


25h 40' 


20.0 ' 13.12 Lost 7.2 per cent. 


17-18 


3.75 




10:05 


9:05 


23h 


21.0 16.79 0.195 | Restless. 


18-19 




9:10 


10:15 


25h5' 


20.5 


17.16 


Somewhat restless. 




: 








1 


Swims about con- 














1 i 


1 sidcrably. 


19-20 






10:20 


10:30 


24h 10' 


21.0 ' 15.77 


Swims around rest- 






1 






! , lewly. Feces. 


20-21 






10:35 


9:40 


23h5' 


21.5 15.43 


Somewhat restless. 














, I Feces. Lost 9.4 per 
















1 cent. 


21-22 


3.66 




9:50 


9:10 


23h 20' 


21.5 


17.23 0.201 Very restless. Feces. 


22-23 




; 9:15 


8:35 


23h 40' 


21.0 ; 14.08 


Swims around r e s t- 








8:40 








lessly. 


23-24 






9:05 


24h 25' 


21.0 


17.95 


1 


Restless. 


24-25 




; 9:10 


8:35 


23h25' 


21.0 


15.16! , 


During early part of 














1 


experiment 5wam 














i 

1 




around much. Lost 












1 




16.83 per cent. 


25-26 


3.36 




&:45 


8:55 


24h 10' 


21.0 16.46 


0.203! 


Same. 


26-27 






9:00 


9:00 


24h 


20.5 


13.37 ! ! Swam around all nfter- 
















noon. 


27-28 






9:05 


9:05 


24h 


20.5 12.26 


1 1 Restless. 


28-29 


\ 9:1(^ 


9:30 

10:C5 

9:35 


24h20' 


20.5 


11.57 1 Lost 20.3 percent. 


29-30 


3.22 




9:40 


24h 25' 


20.5 


14.00 0.178 Rratless. 


30-31 






10:05 


23h30» 


20.0 


15.07 




Restless. 


31-1 






9:40 


10:05 


24h 25' 


20.5 


16.43 


Restless. 


August 














; 1 


1-2 






10:10 
10:30 


10:25 


24h 15' 


21.0 


15.66 


Very rootless. 


2-3 




10:30 


24b 


21.5 


13.44 


1 


Lost 27 per cent. 


3-4 


2.95 




10:35 


10:40 


24h V 


21.0 


11.82 0.167i 




4-5 






10:45 


10:45 


24h 


21.5 


11.09 


Swims about r e s t- 
















lessly. 


5-6 






10:50 


10:50 


24h 


21.0 


12.80 1 HestUss. Lost 29.7 














1 


j 1 per cent. 


6-7 


2.84 




11:00 


10:30' 


23h30' 


21.0 


14.72 


0.220 Restless. 


7-8 






10:35 


11:05 


24h30' 


19.7 


12.18 




Restless. 


8-9 






11:10 


11:50 


24h 40' 


17.5 


9.90 


Very restless towards 


















1 




evening. Quiet next 


















morning. 


9-10 






11:55 


11:55 


24h 


17.5 


8.84 0.139, 0.011 Rather quiet. Lost 






1 

1 1 


1 


p 






! 


33.4 per cent. 


10 


2.69 


34.25 




668h 45' ! 


394.68 









Sergius Morgulis 43 

cc. of oxygen, or an average of about 0.74 cc. per hour. While at 
the start it consumed 1.02 cc. of oxygen per hour, it took only 
0.30 cc. per hour on the last day, or less than one-third. It must 
be pointed out, however, that at first this flounder was more or 
less restless, and during the last two to three days, owing to an 
abrupt change of temperature, its oxygen consumption was greatly 
reduced. 

The second flounder, J5, was smaller, weighing only 4.04 grams, 
and fasted likewise 668 hours, 45 minutes. Its oxygen require- 
ment during the second twenty-four hours after the last feeding was 
0.181 cc. per gram per hour and 0.021 cc. per square centimeter 
per hour. This animal behaved very quietly for the first few days, 
but from the fifth day of the experiment it grew continually more 
restless. It is interesting to record that though the anitnal had no 
food it continued to throw off small quantities of feces until the 
eighth day. It had lost a third of its body-weight (33.4 per cent) 
when it was killed at the end of twenty-eight days. It had used 
up 394.68 cc. of oxygen, or on an average 0.591 cc. per hour. Its 
ox>'gen consumption diminished to one-half, being 0.74 cc. and 
0.39 cc. per hour on the first and last days of the experiment, re- 
spectively. 

From Tables III, IV, and V it will be seen that the first flounder, 
2), lost 1.355 grams of water, 0.353 of a gram of protein, and 0.089 
of a gram of fat in the course of twenty-eight days of fasting. In 
other words, almost one-third of the water content of the body 
had been lost, and about one-half of the protein, while, at the same 
time, nearly nine-tenths of the fat had been wasted. When the 
share of each of these constituents in the total loss is considered, 
it is seen that practically three-fourths of it is water (74.45 per 
cent), while the protein and fat contribute 18.3 and 4.89 per cent, 
respectively. 

Similarly, flounder E had lost 0.992 of a gram of water, 0.300 of 
a gram of protein, and 0.058 of a gram of fat. In the case of this 
specimen, however, more of the protein but less of the fat material 
had been lost. But in both flounders the relative contribution 
of each of these materials to the total loss is about the same. 

With these data it is possible to compute with considerable 
accuracy the amount of oxygen necessary to oxidize completely 
the material lost during the fast. Under the heading ''Oxygen 



44 



Studies on Fasting Flounders 



required** in the last two tables it will be found that for flounder 
D 520.7 cc. of oxygen would have been necessary, and for flounder 
E 406.9 cc, to bum the material utilized in their maintenance metab- 
olism. These computed figures compare very favorably with the 
data of the oxygen consumption actually measured and recorded 

TABLE 111. 
Composition of 100 grams of flounder. 



Normal flounder 

Flounder D, fasting 28 dys 
Flounder Bt fastinc 28 dys. 



WATXB 

80.25 
83.29 
83.04 



PBOTCIN 



FAT 



13.92 
10.41 

9.11 



1.80 
0.83 
0.66 



AHH 

3.38 

T 



UliDETEB- 
MINED 

0.A4 



TABLE IV. 



COMPOSITION 

I ni tial (com puted ) . . . 
Fin.**! (doterminod).. . 

Loaeiu Km 

Loss in per cent 

Uelative loss in per 

cent 

Oxygen required 



Composition of flounder D. 



TOTAL 
WEIGHT 

5.29 gm. 

3.47 gm. 

1.82 gm. 

34.40% 

34.40% 
520.7 cc. 



Carlx>n dioxide 393 .9 cc. 



WATBB ! PBOTBIN 



4.245 gm. 

2.890 gm. 

1.355 gm. 

31.21% 

74.45% 



0.736 gm. 
0.383 gm. 
0.353 gm. 

47.96% 

18. .3% 
341 .Occ. 
266.3 cc. 



PAT 


ASH 


O.lOOgm. 
0.011 gm. 
0.089 Km. 


0.174 Km. 

? 
T 



89% 

4.89% 
179.7 cc. 
127.6 cc. 



UICDETER- 
MIKBD 

0.034 Km. 

T 
T 



TABLE V. 
Composition of flounder E. 



coifposmox 



TOTAL 
WEIGHT 



Initial (computed)... 4.04 Km. 

Final (determined).. . > 2.69 Km. 

Loss in gm ; 1.35 gm. 

Loes in per cent | 33.40% 

Relative loss in per: 



cent 

Oxygen required. 
Carbon dioxide.., 



33.40% 
406.9 cc. 
309. 5oc. 



WATEB 



3.242 gm. 

2.250 gm. 

0.992 gm. 

30.60% 

73.49% 



PBOTEIN 



0.662 gm. 

0.262 gm. 

0.300 gm. 

53.38% 

22 22% 
289.8 cc. 
226.35 cc. 



FAT 



0.076 gm. 

0.018 gm. 

0.058 gm. 

76.31% 

4.30% 
117.100. 
83.15CC. 



ABH 



0.133 gm. 



UNDBTEB- 
MINBD 



0.026 gm. 

? 
? 



in the first two tables. There we found that flounder D con- 
sumed 485.6 cc. of oxygen, and flounder E used up 394.7 cc, or 
6.7 and 3 per cent less than was expected from the amount of 
wasted tissue. It is very significant, and this is a point worth 
emphasizing, that in every case I find the theoretical oxygen re- 



Sergius Morgulis 45 

quirement somewhat higher than the actually determined amount; 
while Putter in his experiments found the reverse. But Putter 
did not accurately determine the oxygen consumption of the 
fasting fish. His method, as was shown above, could not but 
exaggerate the errors. The difference which I found between the 
amount of oxygen used up in a respiration experiment lasting 
practically the full length of the fast, and the oxygen as computed 
from the wasted body material (6.7 and 3 per cent) is within the 
limits of experimental error, considering that in the estimation 
of the initial composition of the flounders, at the moment the 
fast began, slight errors are unavoidable. 

These experiments demonstrate conclusively that aquatic ani- 
mals depend on particulate food, and subsist on their own tissues 
when deprived of food. Organic matter dissolved in sea water 
apparently plays no r61e in the nutrition of these animals. In 
this connection we may recall that Kerb* has shown that aquatic 
animals do not absorb the sugar in solutions in which they live. 
The results here recorded justify the statement that the last sem- 
blance of evidence in favor of Putter's hypothesis is disproved. 

An examination of the composition of the starved flounders 
shows that the absolute loss of protein is much greater than the 
loss of fat; being 4:1 and 5:1 in flounders D and jB, respectively. 
This fact is very interesting, because in the case of fasting mam- 
mals the relation is entirely reversed. Thus in the two famous 
subjects of fasting experiments, Cetti and Breithaupt, the rela- 
tion between the consumed protein and fat was, in round numbers, 
1 :5. In very few instances on record has the protein contributed 
much more to the total loss of material. The predominant part 
which protein plays in the fasting metabolism of fish indicates 
the importance of the nitrogen metabolism of fish as the principal 
source of energy. The total energy expenditure of flounder D 
was 2399 small calories, and that of the smaller fish, £, 1871 cal- 
ories. Of t^Jis amount of energy 64.6 and 61 per cent were derived 
from protein. 

In the last row of Tables IV and V, besides the oxygen re- 
quired to oxidize completely the body materials, the quantities 
of carbon dioxide developed in that process are likewise com- 

• W. Kerb: tJber denN&hrwert der im Wasser gelosten Stoflfe, Internal. 
Rer. d, ges. Hydrobiol, u. Hydrogr., iii, 1910. 



46 Studies on Fasting Flounders 

puled. Dividing these figures by the corresponding amount of 
oxygen, we find that the average respiratory quotient for the whole 
fasting period is 0.756 and 0.761 for flounders D and E, respectively, 
(H It? ^^d 1 1 l\l.) These quotients are rather high, but they like- 
wise point to the important contribution of the prot-eins of the 
body to the metabolic exchange. 



AN EXAMINATION OF THE FOLIN-FARMER METHOD 
FOR THE COLORIMETRIC ESTIMATION OF 

NITROGEN. 

By JOSEPH C. BOCK and STANLEY R. BENEDICT. 

{From the Department of Chemistry of Cornell University Medical College, 

New York City.) 

(Received for publication, November 17, 1914.) 

INTRODUCTORY. 

Many micro methods of analysis have recently been described 
in the literature, which provide for the determination of a great 
variety of substances. These methods open new avenues of ap- 
proach to the study of many problems; but for the most part they 
have been advocated by their authors as not only suitable when 
only small amounts of material are available, but are offered as 
complete substitutes of our standard methods of analysis. 

In order that any method may reach its fuU sphere of useful- 
ness it is necessary that its limitations as well as its advantages 
should be clearly understood. We have had occasion in this 
laboratory to examine in some detail several of the micro pro- 
cedures recently advocated and shall report upon some of these 
methods in subsequent communications. The present paper sum- 
marizes our findings concerning the colorimetric method for total 
nitrogen, as proposed by Folin and Farmer.^ 

The known sources of error in the Folin-Fanner colorimetrie method. 

The Folin-Farmer method has been advocated by its authors 
as fully replacing the ordinary Kjeldahl method. The accuracy 
of the Kjeldahl method is so well known and the amount of work 
and conclusions based upon its use are so vast, that any method 
offered as an alternative to it should be subjected to the closest 
scrutiny. A far closer examination is needed than if the method 

^ O. Folin and C. J. Fanner: this Journal, xi, p. 493, 1912. 

47 



48 Folin-Farmer Method for Estimation of Nitrogen 

were simply offered as available when one mu^t have results at 
once, or when only a small amount of material is available. 

The colorimeter is at the basis of the method we are discussing 
and at the outset it is well to remember that the accuracy of the 
best colorimeter available — ^the DuBoscq — ^is limited to about 1 
per cent of. the quantity of substance employed for the determina- 
tion. 

A possible second 1 per cent error is introduced in the measure- 
ment of 1 cc. of solution employed for a determination. One cc. 
pipettes* are not accurate to within less than 1 per cent when 
most carefully used; under ordinary conditions the error is apt to 
be much larger. 

At the outset, then, we meet two possible 1 per cent errors, 
which the best technique cannot guard against. Either one or 
both of these errors may be apparent in any one determination, 
and may or may not neutralize the other. When it is remem- 
bered that Jbhe Kjeldahl method, as it is ordinarily employed, with 
25 to 50 mgm. of nitrogen for each determination, has an accuracy 
of 0.1 to 0.2 per cent when properly carried out, it will be seen 
that the Folin-Farmer method unquestionably has limitations in the 
direction o* accuracy which do not apply to the Kjeldahl process. 

A fur. her point which we feel has not been suflSciently empha- 
sized by Folin and Farmer and which is even of more importance 
in connection with the work upon blood of Folin and Denis* is the 
question of purity of reagents as regards the freedom from anamo- 
nium salts. When working with Nessler's reagent it is well to 
bear in mind that most substances will give a fair test for am- 
monium with this solution. Furthermore, the quantities of am- 
monia present in some instances are quite sufficient to represent 
1 to 3 per cent of the 1 mgm. which is used in a determination in 
urine. We have found that ordinary c.p. potassium sulphate con- 
tains from 0.005 to 0.01 mgm. of nitrogen per gram, from which 
it can be freed only by recrystallizing twice from pure distilled 
water. 

Folin and collaborators have nowhere mentioned that it is im- 
possible to obtain sulphuric acid free from ammonia. Ordinarily 

' Folin and P'armcr advocate the use of the Ostwald pipette. We have 
obtained better results using the Mohr pipette, graduated in This cc. 
' (). Folin and W. Denis: this Journal, xi, p. 527, 1912. 



J. C. Bock and S. R. Benedict 49 

the c.p. acid contains from 0.01 to 0.03 mgm. of ammonia per 
cubic centimeter, a quantity equivalent to from 1 to 3 per cent of 
the quantity of nitrogen taken for a urine analysis in this method. 

It may be replied that the lack of purity of reagents is not a draw- 
back, since a blank can be run and a correction made. As a mat- 
ter of fact, however, results are apt to be less satisfactory where a 
correction is made for such quantities of ammonia as are indicated 
above, than where such corrections are omitted. This fact is 
probably due to two factors: first, that the accuracy of the method 
is not up to 3 per cent of 1 mgm. of nitrogen; and, second, that 
other errors partially counterbalance the ammonia present in the 
reagents. 

Perhaps one of the most interesting points we have encountered 
in connection with the investigation of the Folin-Farmer and the 
Folin-Denis nitrogen method was in relation to the ammonia in 
air, and to the question of complete absorption of ammonia from 
air by acid solutions. Throughout the work we have employed 
the following technique for the detection and relative estimation 
of small amounts of ammonia. The procedure was to collect the 
ammonia in a few cubic centimeters of dilute hydrochloric acid in 
a test-tube, nesslerize this solution with a measured volume of re- 
agent, and add the same volume of reagent to test-tubes contain- 
ing an equal volume of solution in which were varying amounts 
of ammonia, usually 0.01, 0.02, and 0.03 mgm. of nitrogen,* and 
compare the colors obtained. 

With such technique it is easy to demonstrate that when ordinary 
laboratory air passes rapidly through a single wash-bottle of dilute or 
50 per cent sulphuric acid for ten minutes, from 0.01 to 0.04 mgm. 
of ammonia will escape absorption by the acid. This is even more 
true of the dilute acid used by Folin in the absorption of the am- 
monia to be determined. The contents of a test-tube arranged 
so that the air passes from the Folin absorption flask through 
2 cc. of water containing 1 cc. of ^ hydrochloric acid will show 
that a few hundredths of a milligram of nitrogen have escaped 
absorption. When it is remembered that 0.01 mgm. represents 
1 per cent of the quantity of nitrogen used in a determination, this 
fact is of some importance. It is of far more interest in connection 

* These quantities of ammonia are readily obtained by dilution of the 
standard solution. 

TB« iOUBNAL OF BIOLOOIOAL CBBllfflTmT. VOL. XX. NO. 1 



50 Folin-Farmer Method for Estimation of Nitrogen 

with blood work of Folin and Denis,^ where the total quantity of 
nitrogen estimated frequently did not exceed a few hundredths 
of a milligram. 

A further point in this same connection is the fact that if am- 
monia-free air (obtained by passing the air through three tall wash- 
bottles containing varying concentrations of sulphuric acid) is 
passed rapidly for ten minutes through 1 cc. of tt hydrochloric 
acid and 2 cc. of water, both ammonia-free; contained in an open 
test-tube (or even if the test-tube has a small funnel in the neck), 
the solution at the end of the aeration will, upon being nesslerized, 
be found to contain from 0.005 to 0.02 mgm. of ammonia nitro- 
gen. In this instance the acid solution absorbs ammonia from the 
air of the room due to the spattering during aeration. If the air 
is led into the test-tube through a tube contained in a two-holed 
stopper, with another tube for an outlet, the absorption of am- 
monia from the air in the room is nearly altogether prevented. 

These findings have led us to question the accuracy of the results 
reported by Folin and Denis* in connection with ammonia in blood. 
These writers report figures for ammonia in various bloods out to 
an almost infinitely small amount (0.05 mgm. per 100 cc. of blood), 
yet they report no special precautions to guard against the sources 
of error mentioned above. 

Residts of urine analysis by the colorimetric method as compared 

mith the Kjeldahl tnethod. 

In Table I will be found reported analyses of some seventy 
samples of urine, both by the Folin-Farmer method and by the 
Kjeldahl process. As regards the technique employed, we may 
say that no pains were spared in the carrying out of either pro- 
cedure to obtain the best possible results. The glassware employed 
was all standardized, and the standard acid and alkali for the 
Kjeldahl method were corrected with the greatest exactness. As 
a standard solution in the Folin-Farmer method, we compared 
and frequently used three different solutions. One was prepared 
from ammonium sulphate, obtained as described by Folin,^ one 

» Folin and Denis: this Journal^ xi, p. 161, 1912. 
•Folin and Denis: this Journal, xi, p. 161. 1912. 
'One sample we prepared ourselves; a second was obtained from Dr. 
Emerson of Boston; both had the same value. 



J. C. Bock and S. R. Benedict 51 

from Kahlbaum's ammonium chloride, "zur Analyse," while a 
third was prepared from the urines to be analyzed.* In this latter 
case a measured volume of a sample of urine which had previously 
been analyzed by Kjeldahl's method was digested and then dis- 
tilled into the theoretical quantity of standard acid, and the dis- 
tillate diluted to a definite volume. 

All of these standards had the same color value when nesslerized. 

The air current was obtained from a Crowell pressure blower.^ 
The air current was slow during the first two minutes, and as 
rapid as the apparatus would stand for the last eight minutes. 

A comparison of the results obtained by the Folin-Farmer 
method with those obtained by the Kjeldahl will show that there 
are discrepancies between the two methods of from —11 per cent 
to +3 per cent. The large majority of the results range from 
2 to 4 per cent lower by the FoUn-Farmer method than by the 
Kjeldahl procedure. 

An interesting point in connection with the Folin-Farmer pro- 
cedure is that although one is using seemingly exactly the same 
technique, many determinations will show a difference from the 
Kjeldahl method of only about 2 per cent, while others will show 
a variation of 8 to 10 per cent.*® 

In view of these results it would appear that the Folin-Farmer 
method should not be employed in urine analysis where certain 

* We have not experienced the diflficulties reported by Folin and Farmer 
as regards the obtaining of pure ammonium salts on the market. Kahl- 
baum's ''zur Analyse" ammonium chloride is altogether free from pyridine 
bases or other impurities, at least in detectable quantities. 

* We believe that if the air current is to be employed for the transfer of 
the ammonia, a positive pressure blower is essential in most laboratories. 
We have found numerous laboratories (our own included) where the water 
pressure is not nearly equal to the forty pounds required to operate a suc- 
tion pump satisfactorily for this method. Even where the water pressure 
may be adequate at times errors are apt to occur due to a change in such 
pressure without notice. Compare also F. C. McLean and L. Selling: 
this Journal, xix, p. 36, 1914. 

*J We believe that the large error not infrequently encountered when 
using the Folin-Farmer procedure may be due to the fact that while in many 
determinations the errors, inherent in the method, may neutralize each 
other, sometimes they may all tend in the same direction, thus producing a 
result which is incorrect by from 5 to 10 per cent. 



52 Folin-Farmer Method for Estimation of Nitrogen 

and accurate results are desired. We should, for example, never 
think of using the Folin-Farmer method in making a nitrogen 
balance on an animal. A variation of a gram of nitrogen per day 
by the two methods is not at all uncommon when working on 
human urine. 

Certain modifications of the Folin-Famier method. 

During the work on the Folin-Farmer method two modifications 
of the procedure were developed and the results obtained with 
them are recorded in columns three and four of Table I. In 
Modification No. 1 we have employed distillation by boiling in- 
stead of by the air current. This procedure is simpler and more 
rapid than where the air current is employed and eliminates the 
uncertainties arising from an insufficient air current and from 
failure of complete absorption. The digestion is carried out in 
a test-tube exactly as described by Folin and Farmer, two small 
glass beads being used to prevent bumping. After the mixture 
has partially cooled, 7 cc. of water are added and the tube is 
closed by inserting a two-holed rubber stopper, through one hole 
of which passes a long tube, reaching nearly to the bottom of the 
test-tube, and roughly standardized to hold 3 cc, while through 
the second hole passes a short tube, bent for connecting to a con- 
denser (see Figure 1). 

The long tube should be previously filled with 3 cc. of saturated 
sodium hydroxide solution by suction and closed by means of 
a rubber tip and a pinch-cock. The outlet tube is then con- 
nected with the condenser." A volumetric flask containing 1 to 
2 cc. of T^ hydrochloric acid and enough water to cover the 
connecting tube is used as a receiver. The alkali is allowed to 
run into the test-tube, and the fluids are mixed by blowing a few 

• 

^^ The complete apparatus is shown in Figure 1. A small Licbig con- 
denser which may be easily prepared in the laboratory suffices. A piece of 
glass tubing about 30 mm. by 150 mm. is fitted on each end with a double- 
holed cork or rubber stopper. One long glass tube of about 4-6 mm. inside 
diameter goes through these stoppers and serves as a condensing tube. 
Two short bent glass tubes serve as inlet and outlet for the cooling water. 
The lower en<l of the condenser is connected with a glass tube (or, better, 
an old pipette, to prevent back suction) which reaches into the volumetric 
flask. 



J. C. Bock and S. R. Benedict 

TABI£ 1.' 



<-«» 


"i§" 




Ultt,nm M HODI-iAMALnU DIHODT- 
nCATIOH HO. 1 «r 1 TlCiTION WO. » OF 




^SS^'"^' 


Per »rt of dinpr- Per caol ot diflo> 
■Dogtram Kjsld^l.eucelniin KJEldahl 


1 

2 


4.1B5 
3.372 


3,720 

-10.67 

3.101 


4.020 1 4,020 

-3.47 1 -3.47 

3.323 ! 3.311 



-3.67 

3.470 

+0,99 



3.715 
+3.4S 

3.845 
-0.13 

3.355 
-1.73 

+2.57 



-4.48 


+3.33 


+ 1-12 


3.S45 


3.756 


3-648 


+4.72 


+2,28 


-0 65 


7.580 


7,873 


7,220 


-1.05 


+2.74 


-5.74 


3-225 


3.289 


3.268 


-1.07 


+0.88 


+0.25 


3.289 


3,571 


3.703 


-2.86 


+5.46 


+9.35 


3,333 


3.571 


3.571 


-3-55 


+3.32 


+3 32 


3.448 


3 333 


3.571 


-1,65 


-4.93 


+1.S5 


3-433 


3 448 


3.350 


+0.67 


+ 1.41 


-1 47 


3 423 


3,378 


3.225 


+2.79 


+ 1,44 


-3-15 


3.225 


3.423 


3.448 


-5.53 


+0.26 


+0.90 


3-125 


3.20,i 


3-225 


-3.23 


-0,61 


*0 



* Ttw fiffons rcpnwDt ctmnu ot pitToc™ io 



54 Folin-Farmer Method for Estimation of Nitrogen 



TABLE 1— Continued. 



UHIMV 
SAMPLE 


RESULT 

or ANALYSIS 

BT KJELDAHL 

METHOD 

3.310 


RESULT 

OF ANALYSIS 

BT FOLIN-FARMEB 

METHOD 


RESULT OF 
ANALYSIS BY MODI- 
FICATION NO. 1 OF 
FOUN-FARMER 
METHOD 


RESULT OF 
ANALYSIS BY MODI- 
FICATION NO. 2 OF 
FOLIN-FARMEB 
METHOD 


NO. 


Per cent of differ- 
ence from Kjeldahl 
method 


Per cent of differ- 
ence from Kjeldahl 
method 


Per cent of differ- 
ence from Kjeldahl 
method 


20 


3.125 


3.311 


3.214 






-5.59 


+0.03 


-2.90 


21 


3.287 


3.185 


3.325 


3.325 






-3.10 


+ 1.15 


+1.15 


22 


3.287 




3.311 
+0.72 


3.288 
+0.03 


23 


3.412 


3.325 


3.497 


3.423 






-2.55 


+2.49 


+0.32 


24 


3.412 


3.325 


3.401 


3.268 






-2.55 


-0.32 


-4.22 


25 


2.770 


2.787 


2.778 


2.621 






+0.61 


+0.28 


-5.38 


26 


2.770 


2.703 


2.747 


2.663 






-2.42 


-1.08 


-3.86 


27 


3.084 


3.227 


3.227 


3.011 






+4.63 


+4.63 


-2.37 


28 


2.670 


2.778 


2.825 


2.604 






+4.04 


+5.80 


-2.46 


29 


2.670 


2.710 


2.797 


2.591 






+1.49 


+4.75 


-2.96 


30 


2.888 


2.747 


2.959 


2.874 






-4.87 


+2.45 


-0.48 


31 


2.888 




2.941 
+1.83 


2.924 
+ 1.24 


32 


3.260 


2.959 
-9.24 




3.325 
+ 1.99 


33 


3.260 


3.086 


3.288 


3.185 






-5.31 


+0.85 


-2.30 


34 


3.238 


3.086 


3.145 


2.770 






-4.69 


-2.87 


-14.20 


35 


3.238 


3.125 


3.227 


2.747 






-3.48 


-0.34 


-15.15 


36 


3.421 


3.125 


3.125 


3.325 






-8.65 


-8.65 


-2.80 


37 


3.218 


3.011 


3.125 


3.227 






-6.42 


-2.79 


+0.28 


38 


2.815 


2.747 


2.763 


2.763 






-2.41 


-1.84 


-1.84 


39 


2.866 


2.703 


2.841 


2.762 


• 




-5.67 


-0.87 


-3.62 



J. C. Bock and S. R. Benedict 

TABLE 1— ContintMci. 









BMIF1.TOF 


>Bri.T or 












.^SS, 


■T «jeu>*aL 


""^'^^roT"' 


K)I.rK-^AIU(lH 


"'i™*™'" 




PH(»dtoi<iia.[- 


Par o™t of difl.r- 


P«Mnll>fdlff»,- 








eKefram K]eld>U 


encs Irani Kjeldsbl 






■nothod 


method 


niEtliod 


40 


3.305 


3-200 


3.346 


3.346 






-5.73 


-1.44 


-1.44 


41 


2.530 


2.304 


2.439 


2.424 






-8.93 


-S.JJS 


-4.18 


42 


3.413 


3-347 


3,322 


3.322 






-4,86 


-2.66 


-2.06 


43 


3 141 


^.890 


3.108 


3.125 






-8.00 


-1.11 


-0.50 


44 


3-358 


3.106 


3.423 


3.350 






-7.50 


+1.03 


-0.:'3 


45 


3.333 


3.268 


3,350 


3.325 






-1.62 


+0,81 


+0.06 


46 


3,246 


3.227 


3.225 


3.205 






-0.57 


-0.63 


-1,24 


47 


2.340 


2,294 


2.326 


2 370 






-2.39 


-0 59 


+ 1.28 


48 


2.620 


2.451 


3.591 


2.621 






-6.44 


-1.10 


+0.04 


49 


2.522 


2,500 


2.513 


2,525 






-0,88 


—0,30 


+0,11 


50 


3.871 


3.758 


3,818 


3.846 






-2.02 


-1.37 


-0.64 


51 


3 905 


3.730 


3.875 


3.846 






-4.47 


-0.76 


-1 51 


52 


4.404 


4.168 


4,348 


4.425 






-5.97 


-1.27 


+0.47 


53 


4.384 




4 323 

-1.39 


4.560 

+4.01 


54 


4.447 


4.286 


4 444 


4 477 






-3.62 


-0.07 


+0,68 


55 


4.278 


4.444 


4.253 


4,286 






+3.87 


-0.58 


+0,20 


56 


4.4M 


4.286 


4.478 


4 444 






-2.68 


+ 1.5S 


+0.91 


57 


4.278 


4.189 


4.237 


4.250 






-2.07 


-0,95 


-0.65 


58 


4.130 


4.098 


4,166 


4.121 






-0.77 


+0.87 


-0.22 


59 


4.110 


4.054 


4.i21 


4.121 






-1.36 


+0.26 


+0.26 



56 Folin- Farmer Method for Estimation of Nitrogen 



TABLE 1— Concluded. 







RKATTLT 


RESULT or 


RESULT OT 






OF AVAf^YAIH 


ANALYSIS BT MODI- 


ANALYSIS BT MODI- 


UBTNS 
BAIfPLK 


RESULT 

OP ANALYAI8 

BT KJEKDAHL 

METHOD 


BT POUN-FABMSR 
METHOD 


FICATION NO. 1 OF 

FOUN-PARMER 

METHOD 

Per cent of differ- 
ence from Kjeldahl 
method 


FICATION NO. 2 OF 

FOLIN-FAHMER 

METHOD 


NO. 


Per cent of differ- 

eocefrom KJeldahl 

method 


Per cent of differ- 
ence from Kjeldahl 
method 


60 


4.152 


4.166 


4.18C 


4.166 






+0.33 


+0.88 


+0.33 


61 


3.490 


3.330 


3.600 


3.240 






-4.58 


+3.15 


-7.16 


62 


3.600 


3.270 


3 510 


8.690 






-9.16 


-2.50 


+2.50 


63 


3.590 


3.330 


3.63 


3.750 






-4.45 


-1.94 


+4.44 


64 


3.780 


3.510 


3.780 


3.810 






-7.10 


±0 


+0.78 


«5 


3.860 


3.750 


3.810 


3.900 






-2.85 


-1.29 


+1.03 


66 


3.770 


3.660 


3.780 


3.800 






-2.92 


+0.26 


+3.18 


67 


3.690 


3.450 


3.870 


3.810 






-6.50 


+4.87 


+3.25 


68 


3.030 


2.670 


2.85 


2.850 






-11.88 


-5.94 


-5.94 


69 


3.120 


3.030 


3.180 


3.060 






-2.88 


+ 1.92 


-1.92 


70 


3.170 

1 


3.030 


3.150 


3.060 




! 


-4.41 


-0.02 


-3.47 


71 


14.94 


14.00 


14.86 


14.740 






-6.29 


—0.53 


-1.33 



bubbles of air through the apparatus. The test-tube is then heated 
to vigorous boiling (over a large free flame), the distillation l)eing 
continued until a separation of salts occurs in the test-tul:)e and 
the mixture begins to bump. This distillation requires about two 
minutes. The test-tube is then disconnected from the condenser 
and the latter washed down with a few cubic centimeters of water. 
The liquid in the receiving flask is diluted and nesslerized, as in 
the Folin-Farmer method. 

This procedure is more convenient than the air distillation, 
and an examination of the results by it (Table I) will show that it 
is also more reliable. Still the results obtained do not at all war- 



J. C. Bock and S. R. Benedict 



57 



rant its substitution for the Kjeldahl method in urine analysis. 
We do, however, believe the technique to be of real service in blood 
analysis, as of greater rapidity and accuracy than the air distilla- 
tion and because the volumetric flask used as a receiver may be as 
small as necessity requires, at least to 10 cc.** 




Fig. 1. 



In Table I under the column headed Modification No. 2 are 
recorded analyses made by a direct dilution method, where dis- 
tillation was whoUy dispensed with.'' The technique employed 

" In case a 10 cc. flask is iised, onlj' 6 cc. of water are added to the di- 
gested mixture in the test-tube, and 1 cc. of tJ hydrochloric acid is us»cd in 
the receiver. The washing down of the condenser can be dispensed with. 

" Gulick (this Journal^ xviii, p. 541, 1914) has very recently proposed a 
procedure similar to this one. Our analyses were made a year or two ago 
and we have not had an opportunity to make manj' trials of the procedure 
Gulick suggests. He reports analyses of only one sample of urine by his 
method. 



58 Folin-Farmer Method for Estimation of Nitrogen 

was as follows: A quantity of urine containing about 5 mgm. of 
nitrogen (1 cc. of a slightly diluted sample of urine is usually the 
right amount) is placed in a large test-tube and digested as in the 
Folin-Farmer method, except that no copper sulphate is employed. 
After cooling, the mixture is washed quantitatively into a 50 cc. 
volumetric flask, cooled, and diluted to the mark. Ten cc. (or 
such a volume as will contain 0.7 to 1.5 mgm. of nitrogen) are 
transferred to a 100 cc. volumetric flask and diluted and nessler- 
ized as usual. The standard is 5 cc. of a solution containing 20 
mgm. of ammonia nitrogen, 4 cc. of sulphuric acid, and 4 gm. of 
potassium sulphate in 100 cc. of solution. Both the unknown and 
the standard are nearly neutralized by adding five to six drops of 
concentrated sodium hydroxide solution before being nesslerized." 
The results by this method are ipore satisfactory than by the 
original Folin-Farmer procedure, but are scarcely as good as by 
the distillation method. We feel that the direct method is not 
quite so safe as the distillation by boiling. The latter method 
practically never shows such large variations from the Kjcldahl 
results as are sometimes obtained with the direct dilution method 
or with the original Folin-Farmer procedure. 

SUMMARY AND CONCLUSIONS. 

We may summarize the results reported in this paper as follows: 

1. Certain sources of error in the Folin-Farmer method are 
mentioned. 

2. Analysis of some seventy samples of urine by the Folin- 
Farmer colorimetric method shows that this method usually agrees 
with the Kjcldahl method within about 2 or 3 per cent. For the 
most part the figures by the Folin-Farmer method are appreciably 
lower than by the Kjeldahl process, but variations occur between 
+4 per cent and —11.8 p)er cent of the total quantity of nitrogen 
present. 

3. The Folin-Farmer method is not to be regarded as equiva- 
lent to the ordinary Kjeldahl procedure in accuracy or reliability. 

" The quantities of salts present give no trouble due to development of 
turbidity, but affect the reading somewhat, so that standard and unknown 
should contain approximately the same quantities. 



J. C. Bock and S. R. Benedict 59 

4 . Two modifications of the Folin-Farmer method are presented, 
one of which is recommended as of service in blood analysis or 
where only small quantities of material are available. 

5. The results reported show that the Folin-Farmer method 
can be employed with advantage in any instances where the total 
quantity of nitrogen present is very small (as in the determination 
of the total non-protein nitrogen of the blood), because a high 
percentage error is of much less importance in such analyses. 



A METHOD FOR THE ESTIMATION OF SUGAR IN 
SMALL QUANTITIES OF BLOOD.^ 

By ROBERT C. LEWIS and STANLEY R. BEiNEDICT. 

{From the Chemical Research Laboratory, General Memorial Ilospitalf and 
from the Department of Chemistry, Cornell University Medical 

College, New York City.) 

(Received for publication, November 17, 1914.) 

The quantitative determination of blood sugar is growing more 
and more indispensable, both to the clinician and to the labora- 
tory worker. In the literature we find frequent references to the 
significance of such determinc^tions. Thus Weiland^ has pointed 
out the importance of the blood sugar picture for diagnostic pur- 
poses. On the first page of his book on ''Glycosuria and Dia- 
betes" Allen' says: '*There is a generally recognized clinical need 
of a method for quantitative blood sugar determination which 
shall be quick, accurate, and adapted for use with small blood 
samples." The method proposed in the present paper has, we 
believe, all of these qualifications. 

Two distinct processes are involved in a quantitative determi- 
nation of blood sugar: (1) the removal of blood protein, and (2) 
the determination of the sugar in the protein-free filtrate. In 
the first of these operations lies perhaps the chief difficulty and 
uncertainty in methods of blood sugar analysis, not on account 
of the difficulty of removing the protein, but because of the dan- 
ger of carrying down some sugar at the same time. Many of the 
older methods call for the washing of the precipitated protein 

* A preliminary report of the method described in this paper was made 
before the Society for Experimental Biology and Medicine, Dec. 17, 1913 
(Proc. Soc. Exper. Biol, and Med., xi, p. 57, 1913-14.) 

* \V. Weiland: CentralbL f. d. ges. Physiol, u. Path. d. Stoffw., v, p. 481, 
1910. 

* F. M. Allen: StiLdies concerning Glycosuria and Diabetes, Cambridge, 
1913. 

6i 



62 Estimation of Sugar in Blood 

entirely free from reducing substances, a matter of much difficult}- 
owing to the gelatinous character of the protein mass. In most of 
the newer methods aliquots of the filtrate obtained after precipi- 
tation of the protein are taken for the sugar detennination. In 
such cases it is necessary to show that there is no adsorption of 
sugar by the protein mass. 

In the following brief summary of the literature on blood sugar deter- 
mination we shall consider the two steps in the process separately. Claude 
Bernard,^ who first demonstrated the presence of sugar in normal blood in 
1849, used acetic acid and sodium sulphate to remove the protein. Lis- 
bonne* has recently revived the use of this method. Seegen* employed the 
Schmidt-Mulheim procedure of protein precipitation with ferric acetat-e. 
In the original method of Abeles,^ as well as in the Bang' modification 
where the washing of the protein coagulum was facilitated by the use of 
the centrifuge, the protein was removed by the addition of an alcoholic 
zinc acetate solution. These methods are no longer used. Schenck* took 
advantage of the property of mercuric chloride and hydrochloric acid to 
precipitate protein; while Lupine and Boulud,*** Schondorff," and Bierry 
and Portier'* adopted a somewhat similar method proposed by Patein and 
Dufou; namely, the use of mercuric nitrate. None of the older methods 
of protein precipitation as a preliminary to blood sugar analysis has re- 
ceived more general usage than that of Waymouth Reid,^' as recommended 
by Vosburgh and Richards** in this country. Phosphotungstic acid in 
dilute hydrochloric acid is the protein precipitant in this method. The 
precipitated protein mass is hard and granular and hence may be ground up 
in a mortar and readily washed free from reducing substance. Oppler** 
has made use of phosphotungstic acid and neutral lead acetate to remove 
the protein in making blood sugar estimations. Alcohol alone or in com- 
bination with other methods has received considerable favor (Bang and 



■* Claude Bernard : Mim. de la Soc. bioLj i, p. 121, 1849. 

* M. Lisbonne: Compt, rend, Soc. de biol.j Ixxiv, p. 474, 1913. 

•J. Seegen: Centralhl. /. Physiol. , vi, p. 501, 1892-93. 

^ M. Abeles: Ztschr, /. physiol, Chem., xv, p. 495, 1891. 

» I. Bang: Festschrift fur Olof Hammarsten, Upsala, 1906. 

9 F. Schenck: Arch. f. d. ges. Physiol., Iv, p. 2a3, 1894. 
*° P. Lupine and Boulud: Jour, de physiol. et de path, g^n.j xiii, p. 178, 
1911. 

1* B. Schondorflf: Arch. f. d. ges. Physiol., cxxi, p. 572, 1908. 

*- Bierry and Portier: Compi. rend. Soc. de bioL, Ixvi, p. 577, 1909. 

" E. Waymouth Rcid : Jour. Physiol., xx, p. 316, 1896. 

** C. H. Vosburgh and A. N. Richards: Am. Jour. Physiol. ^ ix, p. 35, 1903. 

»* B. Oppler: Ztschr. f. physiol. ('hem., Ixiv, p. 393, 1910. 



R. C. Lewis and S. R. Benedict 63 

coworkers,*' Tachau,*' etc.). Herzfeld*' has employed metaphosphoric 
acid as a blood protein precipitant. 

Rona and Michaelis^' have introduced a method of protein precipitation 
entirely new to blood sugar determinations; namely, the use of colloids. 
The principle of this method has been adequately explained by Macleod.*^ 
Michaelis and Rona'^ recommended the use of kaolin and colloidal iron, 
showing preference for the latter in papers published soon after their first. 
In the use of colloids an electrolyte is necessary. Rona and Michaelis** 
employed sodium sulphate and Rochelle salt. For whole human blood 
Rona and Takahashi*' have advised the use of monosodium phosphate. 
The method of Michaelis and Rona has been criticized by Oppler;*^ but 
Schirokauer," Moekel and Frank,'* Frank,*' Wilenko," and many others 
have used it extensively with admirable success. Without a doubt, col- 
loidal iron as a blood protein precipitant has received more general use in 
blood sugar analysis than any of the procedures previously mentioned. 

After the protein has been removed from the blood, several methods are 
available for the determination of the sugar in the protein-free filtrate. 
Fermentation by yeast, although possible of application, is seldom em- 
ploj'ed. The polariscopic determination is very accurate when a sufficient 
quantity of blood has been taken, and this method has the support of a 
large number of investigators. Likewise the reduction of solutions of cupric 
salts has been ^^ddely used. The precipitated cuprous oxide may be deter- 
mined gravimetrically as such or as cupric oxide, the methods being so well 
known that a description of them here would be superfluous. In the use 
of copper solutions the degree of reduction, and hence the quantity of 
sugar, may be determined titrimetrically. The method of Bertrand,** in 

*• I. Bang: Der Bluizucker^ Wiesbaden, 1913; Bang: Biochem. Ztschr., 
vii, p. 327, 1908; I. Bang, H. Lyttkens, and J. Sandgren; Zf«cAr. /. physioL 
Chem., Ixv, p. 497, 1910. 

*" H. Tachau: Deuisch. Arch, f, klin. Med., cii, p. 597, 1911. 

'»E. Herzfeld: Ztschr. f. physiol. Chem., Ixxvii, p. 420, 1912. 

*' P. Rona and L. Michaelis: Biochem. Ztschr. ^ vii, p. 329, 1908. 

^^ J. J. R. Macleod: Diabetes: Its Pathological Physiology, International 
Medical Monographs, New York, 1913, p. 28. 

** L. Michaelis and P. Rona: Biochem. Ztschr., viii, p. 356, 1908; xiv, 
p. 476, 1908. 

** Rona and Michaelis: loc. cit. 

" P. Rona and D. Takahashi: Biochem. Ztschr., xxx, p. 99, 1011. 

** B. Oppler: Ztschr. f. physiol. Chem., Ixxv, p. 71, 1911. 

« H. Schirokauer: Bert. klin. Wchnschr., xlix, p. 1783, 1912. 

*• K. "Moekel and E. Frank: Ztschr. f. physiol. Chem., Ixv, p. 323, 1910; 
xli>, p. 85, 1910. 

" E. Frank: Ztschr. f. physiol. Chem., Ixx, p. 129, 1910-11. 

*' G. G. Wilenko: Arch.f. exper. Path. u. Pharmakol., Ixviii, p. 297, 1912. 

"G. Bertrand: Bull. Soc. chim., xxxv, p. 1285, 1906. 



64 Estimation of Sugar in Blood 

which the precipitated cuprous oxide is filtered off, dissolved in a solution 
of ferric sulphate and sulphuric acid, and titrated with potaspiuni perman- 
ganate, has had extensive use. Michaclis'® has recently modified Ber- 
trand's method so as to make it available for the determination of su^slt in 
small amounts of blood. The reduced cuprous oxide of Fehling's solution 
is determined by dissolving it in Bertrand's ferric sulphate-sulphuric acid 
solution and titrating with y^^ potassium permanganate. Several methods 
have been proposed, the basis of which is the determination of copper left 
in solution after reduction. Thus Lehmann's'^ method depends on the 
reaction 

2 CuSO^ -f 4 KI = CuJ, -f 2 K,S04 + Is, 

the liberated iodine being determined by titratioii with thiosulphate. In 
Bang's'* older titration method the unreduced copper is determined by 
titration with hydroxylamine. Flatow" also uses hydroxylaminc in de- 
termining unreduced ammoniacal copper in this way. In Bang's^^ newer 
method, which is microchemical in that only very small quantities of blood 
are necessary for a determination, the unreduced copper is determined by 
titration with yj^ or jJq iodine solution. 

The goal for which every investigator working on new blood sugar meth- 
ods has been striving is to develop a procedure which will require only very 
small amounts of blood. Bang'^ with his microchemical determination, 
and Michaelis'* with his modified Bertrand method have accomplished this; 
but with these exceptions all the procedures above mentioned require at 
least 10 cc. and preferably 25 cc. or more of blood for a single determina- 
tion. Obviously this is a serious disadvantage. The method of Tachau*^ 
requires 5 to 10 cc. of blowl and depends on the reduction of mercuric cyan- 
ide (Knapp's solution) in alkaline solution, the excess of mercuric cyanide 
being determined gravimetrically. Most of the microchemical procedures 
for blood sugar analysis introduced in the last few years have been based on 
color reactions. These may be divided into two classes: (1) titrimetric, 
and (2) colorimetric, meaning a determination with a colorimeter of the 
intijnsity of the color reaction. In the method of Herzfeld'* the protein- 
free blood filtrate is made alkaline and titrated with methylene blue while 
boiling, the blood sugar decolorizing the dye. The decolorization of an 
alkaline safranin solution by dextrose has been used by Hassclbach and 



"L. Michaelis: Biochem. Zlschr., lix, p. 166, 1914. 

" K. B. Lehmann: Arch. f. Hytj., xxx, p. 267, 1897. 

'- Bang: Der BlxiUurkery Wiesbaden, 1913. 

" Flatow: Dewsch. Arch. f. klin. Med., cv, p. 58, 1911-12. 

'* Bang: Der Blutzucker, Wiesbaden, 1913. 

" Bang: Der Blut ucker, Wiesbaden, 1913. 

*• Michaelis: loc. cit. 

3^ Tuchau: Inc. cit. 

"* Herzfeld: loc. cit. 



R. C. Lewis and S. R. Benedict 65 

Lindhard'^ as the basis of a titrimetric method for the determination of 
sugar in urine. 

Of the colorimetric methods, that of Wacker*° depends on the reaction 
of paraphenylhydrazine hydrosulphate and sodium hydroxide with carbo- 
hydrates, the amount of color produced being proportional to the blood 
sugar. Reicher and Stein*' have applied the Molisch reaction— the pro- 
duction of a red color by carbohydrates with a-napthol and sulphuric acid — 
to the quantitative determination of blood sugar. Forschbach and Sever- 
in** have modified the Autenreith-Tesdorpf*' procedure of estimating sugar 
in urine for the quantitative determination of blood sugar. In this method 
Bang's solution is reduced by the sugar contained in a small amount of 
blood, the degree of decolorization being then detennined with a colorime- 
ter. According to Allen,** all of these colorimetric methods tend to give 
high results. 

Shaffer** has recently suggested a procedure for the determination of 
blood sugar in 5 cc. of blood, in which the blood is freed from protein by 
coagulation by heat and the use of colloidal iron, the filtrate being boiled 
with Fehling's solution, and the reduced copper measured colorimetrically, 
or b}' titration as in Bertrand's method. Shaffer reports very low values 
for the sugar content of normal dog's blood by his method, but offers no 
comparative figures obtained by any standard method upon the same sam- 
ples of blood, or indeed upon the blood of any species. Since the prelim- 
inary work upon the method we wish to report involved some studies prac- 
tically duplicating Shaffer's procedure, except that they were carried out 
upon larger volumes of blood, and the results were compared with the 
figures obtained upon the same samples of blood by the Wa3rmouth Reid 
gravimetric method, we shall discuss certain phases of Shaffer's method, 
although we have not tested the method using his exact technique. 

Shaffer precipitates the proteins from the blood by heating, addition of 
acetic acid, and of colloidal iron, and a little sodium sulphate. The filtrate 
is boiled with Fehling's solution, and the precipitated cuprous oxide is 
determined. Shaffer himself raises the question as to w^hether cuprous 
oxide is completely precipitated from such filtrates, and shows that glucose 



*• K. A. Hasselbach and J. Lindhard: Biochem. Zischr.y xxvii, p. 273, 
1910. 

*»L. Wacker: Ztschr. /. physiol. Chetn,, Ixvii, p. 197, 1910. 

** K. Reicher and E. H. Stein: Verhandl. d. Cong. f. inn. Med., xxvii, 
p. 401, 1910; Munchen. med. Wchnschr., Ivii, p. 1032, 1910; Biochem. Ztschr., 
xxxvii, p. 321, 1911. 

** Forschbach and Severin: Arch. f. exper. Path. u. Pharmakol., Ixviii, 
p. 341, 1912. 

** W. Autenrieth and Th. Tesdorpf: Miinchen. med. Wchnschr., Ivii, p. 
1780, 1910. 

** Allen: loc. cit. 

** P. A. Shaffer: Proc. Soc. Biol. Chem.; this Journal j xvii, p. xlii, 
1914; xix, p. 285, 1914. 

THCJOUBKALOFBIOLOOICALCHaMiaTRT. VOL. XX, NO. I 



66 Estimation of Sugar in Blood 

added to ox blood which has* stood about the laboratory for a few da^^s and 
has lost its reducing power, is not quantitatively recovered; i.e., some of 
the cuprous oxide is held in solution under such conditions. But Shaffer 
holds that the power to hold cuprous oxide in solution exhibited by the 
filtrates above mentioned is exhibited only by blood which is not fresh. 
His reason for this conclusion is that the filtrates from fresh blood incubated 
with yeast will not hold any more cuprous oxide in solution than will an 
incubated suspension of yeast in pure solution, though both will hold ap- 
preciable quantities of the oxide in solution. To our minds Shaffer's ex- 
periment in this connection does not settle the point as to whether some 
cuprous oxide may escape precipitation from the filtrates from fresh 
blood. It is conceivable that both yeast or blood protein may leave a max- 
imum of the substance in solution which is the effective agent in preventing 
complete precipitation of the reduced oxide. It seems that a much better 
criterion would have been to compare the results obtained by the Shaffer 
method and by some standard method, where the same sample of fresh 
blood was employed. « 

In this connection we may report two experiments, one made 
upon ox blood which was a few days old, the other made upon 
fresh ox blood. In both cases some glucose was added to the 
whole blood. The procedure was to divide the blood into two 
measured samples, one being analyzed for glucose by the Way- 
mouth Reid method (precipitation with phosphotungstic acid, 
filtration, and subsequent extraction of the precipitate with boil- 
ing water, glucose being estimated in the combined filtrates by 
the Allihn method) ; the other sample of blood being heated to 
complete coagulation with t?tf acetic acid, colloidal iron being 
then added and the glucose determined in the filtrate by direct 
weighing of the cuprous oxide precipitated upon boiling with Feh- 
ling's solution. This latter procedure for freeing the blood from 
protein was described in our previous paper on blood sugar deter- 
mination, and docs not differ materially from Shaffer's procedure 
except that he adds sodium sulphate, in addition to the colloidal 
iron. Sample 1 of ox blood (some days old, to which some glu- 
cose was added) gave 140 mgm. of sugar per 100 grams of blood 
by the Waymouth Reid method, and only 69 mgm. of glucose 
per 100 grams of blood when the cuprous oxide precipitated from 
the heat-colloidal iron filtrate was determined. Sample 2 of ox 
blood (fresh, but to which some glucose had been added) gave 
128 mgm. of glucose per 100 grams of blood by the Wa>inouth 
Reid method, and 84 mgm. of glucose per 100 grams of blood by 



R. C. Lewis and S. R. Benedict 67 

precipitation of the cuprous oxide from the heat-colloidal iron 
filtrate. While these results are not directly applicable to the 
Shaffer method (because we did not use any sodium sulphate, 
since our procedure gave water-clear filtrates without any addi- 
tion of a salt), we feel that they do call attention to the desirability 
of testing Shaffer's procedure by comparison of results obtained 
by it with some standard method upon the same samples of fresh 
blood for any individual species. 

THE NEW METHOD. 

The red color*' obtained by heating a dextrose solution 
with picric acid and sodium carbonate is employed as the 
basis of the proposed method for the determination of blood 
sugar. The blood protein is removed by precipitation with 
picric acid,*^ a method which lends itself to the purpose only 
iDecause of the fact that picric acid is one of the reagents of 
the color reaction and need not be removed from the protein- 
free filtrate.** The method of blood sugar determination as we 
use it in practice is as follows: Two cc.*^ of blood are aspirated 
through a hypodermic needle and piece of rubber tubing into an 
Ostwald pipette, a little powdered potassium oxalate in the tip 
of the pipette preventing clotting. The blood is drawn up a lit- 
tie above the mark and the end of the pipette is closed with the 
finger. After the rubber tubing and needle are disconnected, 
the blood is allowed to flow back to the mark and is discharged at 
once into a 25 cc. volumetric flask containing 5 cc. of water. The 
contents of the flask are shaken to insure thorough mixing and the 
consequent hemolysis of the blood. Then 15 cc. of saturated 
aqueous solution of picric acid are added, as well as a drop or two 
of alcohol to dispel any foam, and the contents of the flask are 
made up to the mark with water and then shaken. After filtra- 

*• The colored derivative formed is probably picramic acid. 

*' In a preliminary report of this method {loc. cit.), we used heat and col- 
loidal iron to remove the protein. Since the early part of 1914 we have been 
lining picric acid and find the method much improved by its use. 

** Folin (this Journal^ xvii, p. 475, 1914) employs picric acid as a precipi- 
tant for protein in blood, tissues, etc., prior to creatine and creatinine de- 
terminations. 

" A smaller quantity of blood may be used if necessary. 



68 Estimation of Sugar in Blood 

tion 8 cc.^® aliquots are measured out into large Jena test-tubes 
for duplicate determinations. Two cc. of saturated picric acid 
solution and exactly 1 cc. of 10 per cent sodium carbonate are 
added (as well as two glass beads and two or three drops of mineral 
oil), and the contents of the flask are evaporated rapidly over a 
direct flame until precipitation occurs. About 3 cc. of water are 
added, the tube is again heated to boiling to dissolve the precipi- 
tate, the contents of the tube are transferred quantitatively to 
a 10 cc.^^ volumetric flask, cooled, made up to the mark, shaken, 
and then filtered through cotton into the colorimeter chamber. 
The color is compared at once with that obtained from 0.64 mgm. 
of dextrose, 5 cc.^^ of saturated picric acid, and 1 cc. of 10 per 
cent sodium carbonate, when evaporated to precipitation over a 
free flame and diluted to 10 cc. as was the unknown, or against the 
picramic acid standard mentioned below. 

Calculation, If the directions as outlined are followed exactly 
the calculation of the sugar present in the unknown blood sample 
is very simple. The original 2 cc. of blood were diluted to 25 cc, 
and of this amount 8 cc. were taken for a determination. In other 
words, the aliquot contained the equivalent of 8/25 x 2 cc, or 
0.64 cc. of blood. The following formula may be used to find the 
blood sugar content: 

reading of standard ^ , 

Mgm. of dextrose m unknown = ;; ; X mgm. of dextrose 

Or: '■^'^™8 °^ unknown j^ standard. 

, , , , reading of standard 

Mgm. of dextrose per cc. of blood = — — : . - - - X 

readmg of unknown 

mgm. of dextrose in standard 
cc. of blood used 

But the amount of dextrose in the standard is 0.64 mgm., and 
the amount of blood used is 0.64 cc. Consequently the second 
fraction equals unity and the equation is simplified. 

^^ When only a small amount of blood is available, satisfactory results 
may be obtained by using a single larger aliquot. 

** In case of hyperglycaemia the final volume of the reaction fluid is made 
25 cc. or 50 cc, and the results are accordingly multiplied by 2.5 or 5.0. 

" It was found that provided the equivalent of 4 cc. of saturated picric 
acid were present, the quantity of this reagent had no influence on the 
amount of color produced. 



R. C. Lewis and S. R. Benedict 69 

^ , , , reading of standard 

Mgm. of dcxtroee per cc. of blood = ; ; . 

reading of unknown 

The per cent of blood sugar is, Of course, 0.1 of the figure thus 
obtained. 

Permanent standard. A solution of picramic acid^ makes a 
verj' satisfactory permanent standard. The color is identical in 
quality with that resulting from the alkaline-picrate-sugar reac- 
tion. Furthermore this solution of picramic acid keeps perfectly. 
Following is the formula of the permanent standard: 

Picramic acid 0.064 gm. 

Sodium carbonate (anhydrous) . 100 gm. 

Water to make 1000 cc. 

Dissolve the picramic acid with the aid of heat in 25 to 50 cc. 
of distilled water which has been made alkaline with the sodium 
carbonate. Cool and dilute to one liter. This solution has the 
same intensity of color as that obtained by the proposed method 
with 0.64 mgm. of sugar when the final volume of the reaction fluid 
is made 10 cc. We have used only two samples of picramic acid, 
and it is quite possible that other samples might give a solution of 
different color intensity. For this reason the picramic acid solu- 
tion should be standardized before being used. 

Accuracy of proposed method. The accuracy and delicacy of 
this color reaction for the determination of blood sugar was shown 
by the analysis of pure sugar solutions (Table I). Dehn and Hart- 
man" have recently published a method of sugar determination, 
the basis of which is the color reaction used by us.^ Their results 

*' Picramic acid was formerly obtainable only in the European market. 
At our suggestion the J. T. Baker Chemical Company, of Phillipsburg, 
New Jersey, are manufacturing picramic acid, and have placed it upon the 
market in this country. We have examined a sample of their product and 
found it wholly satisfactory. 

** W. M. Dehn and F. A. Hartman: Jour. Am. Chem. Soc.j xxxvi, p. 403, 
1914. 

** Attention should be called to the fact that we first reported our blood 
sugar method before the Society for Experimental Biology and Medicine 
on Dec. 17, 1913. The paper of Dehn and Hartman (loc. cit.) appeared in 
February, 1914. These authors made no attempt to use the method for 
blood sugar determinations. We had used the color reaction successfully 
for the determination of pure sugar solutions since the early part of 1913. 



^o Estimation of Sugar in Blood 

TABLE 1. 
AnalysU of pure dextrose solutions by the proposed fnethods. 

DKXTROSE 

Takoi: Found: 

9n0in« tnffin. 

1.00 1.00 

0.90 0.90 

0.80 0.79 

1.00 1.00 

0.90 0.89 

0.80 0.795 

2.00 2.00 

5.00 5.00 

likewise demonstrate the accuracy of the method. Our method 
has an accuracy of from 1 to 2 per cent of the amount of sugar 
present. 

When we first used picrii^ acid we collected the blood directly 
into a saturated picric acid solution, the results obtained agreeing 
closely with those when our heat-colloidal iron method of protein 
removal was employed. We soon found, however, that, if hemol- 
ysis occurred previous to precipitation of the blood protein, higher 
figures resulted. Table II shows comparable determinations of 
blood sugar by the different modifications of the colorimetric 
method just mentioned. 

The consistently higher figures obtained by the method as de- 
scribed in this paper (precipitation of the blood protein with pic- 
ric acid after hemolysis) should be not^d. Evidently unless hemol- 

TABLE II. 

Comparison of the heat-colloidal iron and picric acid methods of protein 

precipitation as a preliminary to the determination of 

blood sugar by the new colorimetric method. 

The figures give dextrose per cent. 

hi:atkk>lloidal ,ROK I p,eH,cAcn> ! noBio acid 

(blood COLLSCTBD DIBECTLY (XO HBMOLYaiS) | (HEMOLT8I8 PREVIOU8 TO PRE- 

INTO HOT y?^ ACETIC ACID) 



CIPITATION OF BLOOD PROTEIN) 



0.164 0.166 

0.119 0.118 

0.105 0.105 

0.361 0.361 

0.094 0.101 

0.124 0.130 

0.153 0.163 



R. C. Lewis and S. R. Benedict 



71 



TABLE 111. 



Comparison of the gravimetric method of Waymouth Reid for the determina" 
tion of blood ettgar with the new colorimetric method. 







DEXTBOBB 




BLOOD 


GrBvimet- 

ric method 

of Reid 


Colorimettic method (heat-ooUoidal 
iron precipitation)* 


Colorimetric 
method (picric 
acid precipi- 
tation after 
hemolysis) 


Ox 


percent 

0.0588 

0.0596 

0.0985 

0.119 

0.122 

0.111 


per cent 

0.0552 [corrected 0.0585] 
0.0563 [corrected 0.0597] 


per cent 


Ox 




HiiTnan 


0.098 


HnmRTi 


0.119 


Human 


0.122 


Human 


0.108 







*ABwehav«already pointed out, this method of blood protein precipitation yields results 
5 to 7 i>er cent too low (cf . Table 11). Consequently we have made [in brackets] a 6 per cent 
correction. 



ysis tAkes place previous to precipitation of the protein, some 
sugar is lost by inclusion in the red blood cells. 

Table III gives a comparison of blood sugar determinations 
by the proposed method and by the gravimetric method of Way- 
mouth Reid. A very close agreement in the results obtained by 
the two methods is seen.** 

TABLE IV. 
Showing the complete recovery by the new method of dextrose added to blood. 



Ox blood I (orifdnal analysis) 

Ox blood I -h 0.5 ingm. dextrose per cc 

Ox blood I -h 1 mgm. dextrose per cc 

Ox blood II (original analysis) 

Ox blood II + 1 mgm. dextrose per cc 

Ox blood (-h dextrose) III (original analysis) 

Ox blood (+ dextrose) III + 0.5 mgm. dextrose per cc. 
Ox blood (+ dextrose) III + 1 mgm. dextrose per cc. . . 



DEXTROSE 



Theoretical ; Found 



per cent 

0.1043 
0.1543 

0.1710 

0.2080 
0.2580 



per cent 

0.0543 
0.1048 
0.1548 
0.0710 
0.1720 
0.1580 
0.2080 
0.2580 



** When our method was first devised, we feared that it might give slightly 
too high values for blood sugar, owing to the presence in blood of creatinine, 
or of other compounds giving the same color reaction as that given by the 



72 Estimation of Sugar in Blood 

' When dextrose was added to blood it could be quantitatively 
recovered by the new colorimetric method as is shown in Table IV. 
Normal blood sugar content. An examination of the bloods 
of healthy persons has shown a normal blood sugar content of 
0.09 to 0.11 per cent (average = 0.1 per cent). This figure agrees 
with that reported in the literature by the majority of authors. 

glucose. The results do not, however, show any tendency to be higher than 
those obtained by other standard methods. Hence we may infer that the 
traces of creatinine present in blood are not sufficient to afifect the color 
values obtained. 



THE SYNTHESIS OF HIPPURIC ACH) m EXPERIMENTAL 
TARTRATE NEPHRITIS IN THE RABBIT.^ 

By F. B. KINGSBURY and E. T. BELL. 

(From the Departments of Physiology and Pathology of the Medical School 

of the University of Minnesota^ Minneapolis,) 

(Received for publication, November 19, 1914.) 

The object of this investigation was to determine whether ne- 
pliritis influences the synthesis of hippuric acid in the rabbit. 

The chief conditions affecting the synthesis of hippuric acid are known 
from the work of Bunge and Schmiedeberg* and the later experiments of 
Hoffmann,* \V. Koch,* and Bashford and Cramer.* The place of hippuric 
aoid synthesis in the dog is definitely believed to be the kidney, but in the 
case of the rabbit the evidence is somewhat conflicting. 

Salomon,* using the analytical method of Bunge and Schmiedeberg, 
showed that the removal of the kidneys of the rabbit did not prevent the 
synthesis of hippuric acid. The necessary glycocoU and benzoic acid were 
introduced into the stomach of the rabbit after the kidneys had been extir- 
pated, and hippuric acid was found in the blood, liver, and muscles. He 
found that tying off the ureters did not cause any increase of hippuric acid 
in the blood, muscles, or liver over that found when the kidneys were extir- 
pated. Salomon concluded that the formation of hippuric acid in the rab- 
bit occurred in other places than the kidney. 

Jaarsveld and Stokvis^ extirpated one kidney of a rabbit, tied off the 
other at the hilus for 1 hour and 20 minutes, and then introduced 1 gm. of 
benzoic acid into the stomach. The rabbit died in 22 hours, and hippuric 
acid could be found neither in the blood nor in the urine. The same result 



^ Aided by a grant from the Research Fund ol the University of Minne- 
sota. 

* G. Bunge and O. Schmiedeberg: Arch. f. exp. Path. u. PharmakoL, vi, 
p. 233, 1876. 

* A. HofiFmann: Arch. f. exp. Path. u. PharmakoL, vii, p. 233, 1877. 

* \V. Koch: Arch. f. d. ges. Physiol., xx, p. 64, 1879. 

* E. F. Bashford and W. Cramer: Ztschr. f. physioL Chem., xxxv, p. 324, 
1902. 

« W. Salomon: Ztschr. f. physiol. Chem., iii, p. 365, 1879. 
^ G. J. Jaarsveld and B. J. Stokvis: Arch. f. exp. Path, u. PharmakoL, 
X. p. 268. 1878. 

73 



74 Synthesis of Hippuric Acid 

was obtained when both kidneys were removed; no hippuric acid could be 
found in the blood. 

A rabbit, which had previously been fed benzoic acid, was injected sub- 
cutaneously with a 50 per cent glycerine solution, which produced a marked 
hemoglobinuria. The urine was analyzed for hippuric and benzoic' acids 
to see what effect the injury to the kidneys so produced would have on 
the synthesis. In most of the experiments of this kind the percentage 
of the total benzoic acid in the form of hippuric acid was much less than 
the percentage of the free benzoic acid, but in one experiment it was much 
more. In the case of normal rabbits Jaarsveld and Stokvis found nearly 
all of the ingested benzoic acid in the form of hippuric acid in the urine ; 
but in all experiments except one they were able to find weighable quanti- 
ties of free benzoic acid. The glycerine injections prevented the synthesis 
to a large extent. 

A rabbit, which had been given benzoic acid by stomach tube, 'was killed 
5} hours later by bleeding. 0.115 gm. of free benzoic acid and 0.058 gm. of 
hippuric acid were found in the stomach and small intestine. 0.135 gm. 
of free benzoic acid and no hippuric acid were found in the blood, 0.016 gm. 
of free benzoic acid and traces of hippuric acid in the urine, and neither free 
benzoic acid nor hippuric acid in the large intestine. 

The kidneys of a normal rabbit were extirpated, and 1 gm. of benzoic 
acid was introduced into the stomach. The contents of the stomach and 
small intestine were analyzed a short time after and 0.114 gm. of free benzoic 
acid and 0.082 gm. of hippuric acid were found. 

Jaarsveld and Stokvis concluded that in the rabbit the synthesis takes 
place in more than one part of the body; namely, in the kidneys, liver, and 
small intestine. 

Jaarsveld and Stokvis used Bunge and Schmiedeberg's anal3rtical method 
for blood and tissue analysis, and their own modification of this method 
for the urine analysis. 

It is doubtful, however, whether Jaarsveld and Stokvis really produced 
a severe tubular nephritis by the injection of glycerine. They described 
a marked hemoglobinuria in every case. Potter and one of us (Bell) have 
shown that hemoglobin is eliminated by the convoluted tubules of the rab- 
bit kidney, and that when these structures are all destroyed or severely 
inj ured only a very little hemoglobulin is excreted in the urine. The marked 
hemoglobinuria produced by glycerine does not signify the presence of 
severe injury to the kidney. It is due to the liberation of a large amount 
of hemoglobulin in the circulating blood. 

Weyl and Anrep' found that, in rabbits in which fever had been produced 
by the injection of pus, the ability to synthesize hippuric acid was greatly 
diminished. With dogs this was also true, but to a much less extent. 

Van de Velde and Stokvis'^ found that after the injection or feeding of 
benzoato to rab})it8 all tho benzoic acid of the urine was present as hippiiric 

' Th. Weyl and B. von Anrop: Ztschr. /. physiol. Chew., iv., p. 169, 1880. 
' A. van de Velde and B. J. Stokvis: Arch. f. cxp. Path. u. PhnnmikoL, 
xvii, p. 189, 1883. 



F. B. Kingsbury and E. T. Bell 75 

acid, provided the urine was acid. If the urine was alkaline, as much as 
50 per cent might be present as free benzoic acid. These authors called 
attention to the fact that hippuric acid was very easily hydrolyzed in alka- 
line media. This fact was apparently not known to many of the later in- 
vestigators, for in many of the later methods for hippuric acid estimation 
the evaporation of the liquid containing it with alkali is recommended. 

We have found that, after the injection of sodium benzoate, rabbit urine 
may contain large amounts of unconjugated or free benzoic acid (as the Na 
salt if the urine is alkaline, as it usually is) . Other investigators have also 
found this to be the case (Wiener," van de Velde and Stokvis, Jaarsveld 
and Stokvis). We call attention to this fact because it has been assumed 
from time to time that rabbit urine under these conditions contained so 
little benzoic acid, if any at all, that it was not necessary to consider it. 

Howard B. Lewis," in a study of the relation of the hippuric acid nitro- 
gen to urea in rabbit urine, neglected to take account of the free benzoic 
acid present in the urines which he analyzed for hippuric acid on the as- 
sumption that it was not there ; so that his figures for hippuric acid and the 
nitrogen present as hippuric acid cannot be accepted, since it is quite pos- 
sible that a large part of the so called hippuric acid (total benzoic acid) 
might have been present as free benzoic acid. 

METHODS. 

The experimental work of Underbill, Wells, and Goldschmidt" has given 
OS a method for the production of a severe tubular nephritis with a minimum 
of glomerular injury. The nephritis was produced in rabbits by the subcu- 
taneous injection of small amount^ of racemic tartaric acid. 

This is the procedure which we have used in our experiments to produce 
nephritis. By administering large amounts of water by stomach tube we 
have found that it is often possible to get considerable excretion of urine 
from severely injured kidneys. If the animals are not forcibly watered the 
amount of urine obtained is usually too small for accurate analysis. 

The physiologic condition of the kidneys was determined by the phenol- 
sulphonephthalein test introduced by Rowntree and Geraghty.** Rown- 
tree and Fitz** have produced abundant clinical evidence that this test is 
an accurate index of the functional capacity of the kidney. A very low 
phthalein excretion was always associated with a severe clinical case of 
nephritis. Their work has been corroborated by a large number of ob- 
servers. 



»o H. Wiener: Arch. /. exp. Path. u. Pharmakol., xl, p. 313, 1898. 

^^ H. B. Lewis: this Journaly xvii, p. 505, 1914. 

»2 F. P. Underbill, H. G. Wells, and S. Goldschmidt: Jour. Exper. Med., 
xviii, p. 322, 1913. 

" L. G. Rowntree and J. T. Geraghty : Jour. Pharyn. and Exper. Therap., 
i, p. 579, 1909-10. 

1* L. G. Rowntree and R. Fitz: Arch. Im. Med., xi, p. 258, 1913. 



76 Synthesis of Hippuric Acid 

In rabbits Potter and one of us (Bell)^* found, by killing the animals at 
the end of the test and examining the kidneys, that a trace or a zero phthal- 
ein excretion for 2 hours is associated invariably with a severe tubular in- 
jury, and that marked granular degeneration is present when the phthalein 
excretion is as low as 5 or 6 per cent. 

The work of Rowntree and Geraghty, as well as that of Potter and Bell, 
has proved that phenolsulphonephthalein is excreted by the tubules and not 
by the glomeruli. A decrease of phthalein excretion corresponds to tubu- 
lar injury. 

In addition to the functional test with phthalein, two of our animals were 
killed at the end of the experiment, and the gross and microscopic changes 
in the kidneys were studied. The urine was taken from the collecting jar 
at 24 hour intervals. After the injection of tartrate it was collected by 
catheter, so as to separate the 24 hour specimens sharply. The use of the 
catheter is the more necessary since retention of urine is common in the 
nephritic animal. 

A preliminary test with phenolsulphonephthalein was made at the be- 
ginning of the experiment. A known amount of sodiimi benzoate was in- 
jected subcutaneously daily and the urine analyzed for a few days before 
the tartrate was given. It was found that without the subcutaneous in- 
jection of benzoate there was a larger error in estimating the benzoic and 
hippuric fractions of the small amount of total benzoic acid normally ex- 
creted by the rabbit. The rabbits were then injected intramuscularly with 
0.3 to 0.7 gm. of racemic tartaric acid dissolved in 10 to 20 cc. of water 
and neutralized with sodium carbonate. After several hours the phthalein 
test was made. If the phthalein excretion was very low or suppressed, the 
daily injections of sodium benzoate were continued and the urine was col- 
lected by catheter at 24 hour intervals. The animals were watered by stom- 
ach tube, 50 to 100 cc. daily. If not forcibly watered the rabbit will usually 
drink very little, and the amount of urine passed may be too small for sat- 
isfactory analysis. The benzoate also acts as a diuretic. Considerably 
more urine is usually obtained as a result of its use. One must wait several 
hours after the tartrate injection before benzoate is administered, since the 
latter seems to prevent the development of a severe nephritis if injected too 
soon. 

The rabbits were fed on carrots exclusively. They were allowed all 
they would cat of this food. A dish of water was kept in the cage con- 
stantly. Each day about 0.5 cc. of a thymolchloroform solution was put 
into the collecting jar for the urine as an antiseptic. During the earlier 
part of the work enough dilute hydrochloric acid was put into the jar to 
more than neutralize the alkali of the urine; but this procedure was 
soon discontinued since it seemed unnecessary. The urines were analyzed 
immediately so that there was no appreciable loss of ammonia. 

We have found that the Folin and Flanders" method for the determina- 

»* A. C. Potter and E. T. Bell: Am. Jour, Med, Sc, 1915, (in press). 
"O. Folin and F. F. Flanders: this Journal, xi, p. 257, 1912. 



F. B. Kingsbury and E. T. Bell 77 

tion of hippuric acid was very satisfactory, and it is the method that we have 
used for all our analyses. 

Folin and Flanders" were able to extract benzoic acid quantitatively from 
cranberries and tomato ketchup, and to determine accurately the amount 
of this acid in the extract by titration against tenth normal sodium ethylate, 
as they determined the liberated benzoic acid in their hippuric acid method. 
The method that we have used for free benzoic acid is based on the above 
method, and is as follows: 

Method for the determination of free benzoic acid in rabbit urines. 25 or 
.50 cc. of the rabbit urine are transferred by pipette into a small beaker, a 
drop of dilute alizarine red indicator is added, showing the usual alkalinity 
of the urine due to the presence of sodium carbonate, and concentrated hy- 
drochloric acid added from a burette drop by drop until the solution becomes 
acid. The solution is now boiled for a minute to remove the excess of car- 
bon dioxide, as this has been found to interfere materially w^ith the accuracy 
of the method. 

This boiling must not be continued or an appreciable amount of hippuric 
acid will be hydrolyzed and determined with the benzoic acid; but, as will 
be shown, boiling for this short time will not split a measurable quantity of • 
hippuric acid. The cool solution is now transferred to a separatory funnel 
by means of enough water to bring the total volume up to 100 cc, 0.5 cc. of 
concentrated hydrochloric acid is added to make the solution distinctly acid, 
and the solution saturated with ammoniimi sulphate by dissolving it in 55 
gm. of the latter. 

The saturated liquid is now extracted with neutral, alcohol-free chloro- 
form in one50cc., one 35 cc, and two 25 cc. portions, as in the Folin-Flanders 
method for hippuric acid. The chloroform extract is washed with three 
100 cc. portions of saturated sodium chloride solution containing 0.5 cc of 
hydrochloric acid in 1000 cc. of solution. This washing is carried out in a 
second separatory funnel. The washed extract is titrated against tenth 
normal sodium ethylate with phenolphthalein as indicator. Duplicate 
determinations will give titration figures varying by not more than 0.05 cc, 
if carefully made. 

Special experiments showed that the removal of liberated carbon dioxide 
from the acidified rabbit urine before extraction with chloroform is abso- 
lutely essential. 

The following experiment shows that under the prescribed conditions 
only a negligible hydrolysis of hippuric acid occurs. 0.15 gm. of crystalline 
hippuric acid was dissolved in about 25 cc. of hot water, one drop of alizarine 
red indicator added, and the solution made distinctly acid by the addition 
of two drops of concentrated hydrochloric acid. The solution was boiled 
for a minute and allowed to cool, when the hippuric acid separated out as 
crystals. On extracting and titrating the extract, two drops of tenth nor- 
mal ethylate were more than sufficient to neutralize all the benzoic acid that 
had been split from the hippuric acid. 



" Folin and Flanders: Jour. Am. Chem. Soc, xxxiii, p. 1622, 1911. 



78 



Synthesis of Hippuric Acid 



To prove that benzoic acid could be added to rabbit urine and quantita- 
tively recovered by this method, 5 cc of a solution of sodium benzoate hav- 
ing an acid equivalent of 12.25 cc. of tenth normal ethylate were added to 
25 cc. of rabbit urine and the urine was analyzed. 

25 cc. of rabbit urine contained benzoic acid, the equivalent of 2.70 cc. 
-fir ethylate. 

25 cc. of rabbit urine plus 5 cc. of sodium benzoate contained benzoic acid, 
the equivalent of 14.90 cc. -^ ethylate. 

Sodium benzoate by difference, 12.20 cc. -nr ethylate. 

Sodium benzoate by analysis, 12.25 cc. jjr ethylate. 

The difference of 0.05 cc. is negligible. 

Rabbit F, The animal was injected every day at 11 a.m. with 20 cc. of 
a sodium benzoate solution which contained the equivalent of 0.630 gm. of 

TABLE 1. 
Rabbit F. 



V) 



S 

bk 
O 

H 

a 



cc. 



s 






1 

o 

3: 

4 
5 
6l 
/ 

8 



71.0 
115.0 

95.0 
105.0 
115.0 
208.0 
135.0 
157.0 



9' 29.0 



10 108.0 



11 



160.0 



12 187.0 



• 


1 


u 


•< 


z 


1 ► 


u 


1% 




! H 
Q 


I Igi M 


U 


1 , » 


< 


o •< 


P4 


Z X 


, O 


-£ 




s £ 


. K 


0. 


BB 



Feb. 2, 
11 a.m., 
0.4 gm. 
Feb. 3, 
5 p.m., 
0.0 gm. 



z 

u 

a 

a a 
2" 

z < 
wo 

•J M 



z 

c 

§ 






gm. 



Xonnal 




• 





Normal 





630 
630l 
630. 
e30| 
630; 
630 
630| 
OSOj 



^ 



o 

M 

z 

n 
a 

H 



gm. 

1.107 

1.22a 

0.702 
1.246 
0.829 

1 . 167; 

0.556 
0.677 

I 



gm. 

0.076 
0.045| 
O.O33I 
0.042 
0.046 
0.093 
0.131- 
0.055 



5? 



gm. 

0.559 

1.043: 

0.834 

0.834 

0.817, 

0.951 

0.736] 

0.959 



H 

z 

o 

a 

gm. 

0.044 
0.082 
0.065 
0.065 
0.004 
0.074! 



SB Z 

So 

gas 55 

r M H 

S 3 O 



k 






ft- :; 
5^ 



per cent per cent 

3.9 , 83.5 



6.7 
9.3 
5.2 
7.2 
6.4 



0.058' 10.3 
0.075 11.1 



94.3 
94.5 
93.1 
92.4 
87.4 
79.3 
92.3 



Fol). 4, 
less than 



2% 



Feb. 5, 
loss than 



0.630, 0.088 0.015 0.284 0.022 25.3 93.0 



0.630 0.442 0.049; 0.541i 0.042 9.0 8S.4 



0.630 0.999 0.151 0.886 0.069 6.9 80.1 

i , I 






Fob. 7, 0.030 1.014 0.415 0.701 0.055 



3.4 53 . 5 






I 



F. B. Kingsbury and E. T. Bell 



benioic acid, as determined by analysis. On the day that the tartaric acid 
was given, the injection of the sodium bencoate was postponed until 4 p.m. 
It had been found that, if the benioate was given at the same time as the 
tartaric acid, the latter wae not so efficient in its action on the kidneys. 
The resultaof the analyses and of the phthalein tests are recorded in Table I. 

At autopsy the kidneys were considerably enlarged and cloudy in appear- 
once. Crushed pieces of fresh tissue shoned all the tubules to be very gran- 
ular. Sections in parafBn showed an irregularly granular fragmented cyto- 
plasm. No necrosis was present. There wasvome fat in the tubules. 

The percentage of the total benzoic acid present in the urine as hippuric 
acid remained remarkably constant throughout the experiment. In the 
period before the injection of tartaric acid the average percentage of hip- 
puric acid was 89.7 per rent as compared with 78.7 per cent during the period 
when the phthalein test indicated severely damaged kidneys. 

It is interesting to note that the total amount of bensoic acid recovered 
in 12 days was 7.296 gm. as compared with 7.560 gm, injected, althaugh a 
small amount of urine was lost on February 4. On this day the total 
amount of benzoic acid recovered was approximately a third of the average 
normal quantity, and the hippuric acid nitrogen was 25 per cent of the total 
nitrogen. 

TABLE II. 



2 270.0 

3 277.5] 

4 364. OJ 

5 268.01 Feb. U. 

injection 
6'Xone • 

7 80.0 

I 



! 

9. 341 .oj 



Complete 

auppres- 



Complete 
auppres- 



Rabhit G. 



0.630 0.S03 IGft 0,328 



0.630] O.nOOJ 0,S72J 0,324 



)8l 4.1 64.0 

»| 1.7 I 28, G 

0,628J 0,053^ 3.6 j 7:t 3 

0,028J 2,4 i 3S.0 

0,371 0,0291 3.0 | 35 S 



8o Synthesis of Hippuric Acid 

Rabbit G. The conditions of this experiment were similar to those of 
the preceding experiment. The results are recorded in Table II. 

On the seventh and eighth days there was complete phthalein suppres- 
sion, and there was a trace of albumin in the urine of these days. This was 
not removed from the samples of urine that were analyzed for hippuric acid, 
although if present in any quantity it should be removed ; for, on treating 
albumin as is required by the hippuric acid method, a certain quantity of 
benzoic acid can be formed from it, as is shown by the following experiment. 

5 gm. of Merck's egg albumin were boiled 4.5 hours with 25 cc. of con- 
centrated nitric acid and 2Jcc. of water in the presence of 0.2 gm. of cupric 
nitrate. The chloroform extract of this reaction mixture required for neu- 
tralization 18.40 cc. of tenth normal sodium ethylate, equivalent to 0.2294 
gm. of benzoic acid. The acid was identified as benzoic acidi 

It seems to be characteristic of rabbit urines after tartaric acid 
injections that they are thin and watery and usually contain slight 
traces of albumin. In the experiments with Rabbit G there was 
so little albumin present that we are certain that it could not have 
appreciably altered the figures for benzoic and hippuric acids. 

Rabbit G gradually recovered, although there was complete 
suppression of phthalein excretion for two daj^. 

This experiment agrees with that on Rabbit F, that severe 
renal injury did not cause any significant change in the relations 
between the free and combined benzoic acids present in the urine.- 
The average percentage of the total benzoic acid present as hip- 
puric acid for the period just before the injection of tartaric acid 
(February 10 to February 15) was 47.5 per cent, and for the period 
during which there were two days of complete suppression of the 
phthalein the percentage was 38.2 per cent (February 15 to the 
end of the experiment). 

Of the 5.670 grams of benzoic acid injected during the 9 days 
4.348 grams were recovered. 

Rabbit H. The experiment with this animal was conducted in the same 
way as the preceding, except that there was a shorter period before the in- 
jection of tartaric acid. On the fifth day the animal was killed. 

The kidneys were found greatly enlarged and cloudy. Crushed pieces 
of fresh tissue showed marked granular degeneration and considerable fat. 
Paraffin sections showed necrosis of the zone of convoluted tubules adjacent 
to the capsule. The tubules in other situations showed pronounced granular 
degeneration but no necrosis. The glomeruli were apparently not injured. 
This represents therefore a severe tubular nephritis. 

The average percentage of the total benzoic acid present as hippuric acid 
for the 2 days previous to the tartaric acid injections was 77.8 per cent ; and 



F. B. Kingsbury and E. T. Bell 8i 

TABLE m. 
Rabbit H. 

I 



^' I is| I i 



I 0.63» 0.624 



"3 



I.I18.0| : 

2 108.0 Feb. 25. 



injection | i 

3' 29. l' Trace 0.630J 0.087 

4 118,0 |Complctc' 0-630 0.472 

I auppree-' ' 

5 238.0| Complete! 0. 630 1.237 

I suppres- 1 



Si I i 

i'i [ s| 

r ! P i ?i 

s a aS [ i: 



0,827 0-065 5.2 



for the 3 days aft«r the injections, during 2 of which there waa complete 
suppression of the phthalein, the average percentage waa ©7.8 per cent. 
This experiment was also in complete agreement with the preceding ones. 
cid injected during the 5 days, 2.578 gm. were 



Sept. 22, 
4.30 p.m., 
0.5 gm. 



T" 


If 


i 


11 


|g 




em. 


Iw«-,(' 


O.Oi'i 


5.8 


O.OS^ 


6-1 


0,144 


90 


0,025 


12.7 









82 Synthesis of Hippuric Acid 

Rabbit K. The diet of this rabbit consisted entirely of milk, of which 
it was given all that it would drink. The results are recorded in Table IV. 

The percentage of hippuric acid of the total benzoic acid shows a steady 
increase from 32.9 to 91.7 per cent. On the day when there was practically 
complete suppression of phthalein the percentage of hippuric acid was the 
highest, 91.7 per cent. 

This experiment also shows that tartrate nephritis does not alter the 
mechanism of hippuric acid synthesis. 

DISCUSSION AND SUMMARY. 

Our experiments include cases of severe acute nephritis (Rabbits 
G and H) as well as cases of moderate intensity. In Rabbit H 
the functional t€st was checked by the autopsy findings. There 
is no doubt that a complete suppression of phenolsulphonephthal- 
ein is always associated with severe injury to the convoluted tu- 
bules. Reference to the accompanying tables will show, however, 
that there was never any definite interference with the hippuric 
acid synthesis during the nephritis. We must therefore con- 
clude that severe injury of the convoluted tubules does not affect 
the synthesis of hippuric acid. 

The glomeruli and the collecting tubules are not seriously in- 
jured in tartrate nephritis. If hippuric acid is synthesized in the 
kidney of the rabbit, this synthesis must be accomplished under 
the influence of cither the degenerating cells of the convoluted 
tubules, or the cells of the glomeruli or collecting tubules. We 
are inclined to the view that the synthesis of hippuric acid does 
not occur in the kidnev of the rabbit. 



THE SOLUBLE POLTSACCHASIDES OF LOWER FUNGI. 

m. THE INFLUENCE OF AUTOLYSIS ON THE BCYCODEXTRAN 

CONTENT OF ASPERGILLUS NIGER. 

By ARTHUR W. DOX. 
(From the Chemical Section of the Iowa Agricultural Experiment Station.) 

(Received for publication, November 24, 1914.) 

The first two papers^ of this series describe the preparation and 
properties of the two new carbohydrates, mycodextran and my- 
cogalactan, from lower fungi. The material used in preparing 
these substances was the dried fungus from cultures that had just 
reached maturity but had not undergone autolysis. It was stated 
that attempts to prepare these polysaccharides from cultures that 
had autolyzed for six weeks were unsuccessful. The present work 
was undertaken with the view of determining the quantitative 
yield of mycodextran from cultures grown under the same condi- 
tions for different periods of time, and of following the variations 
due to autolysis. 

A large number of liter Erlenmeyer flasks, each containing 
200 cc. of Raulin's medium, were sterilized and inoculated with 
spores of Aspergillus niger. At different stages of growth or autol- 
>'sis, indicated in the table below, the fungus was removed from 
five flasks, washed with distilled water, and dried collectively at 
100°. After drying, the mold was ground to a fine powder and 
extracted three times for a half hour with successive portions of 
100 cc. of boiling water. A clear filtrate was obtained in each 
case, from which the mycodextran generally separated out in a 
white flocculent form. Three extractions were considered sufficient 
since a fourth extraction gave only a trace of mycodextran. The 
crude mycodextran was filtered with suction on a hardened filter, 
washed with cold water until the color had been removed, then 

'A. W. Dox and R. E. Neidig: this Journal, xviii, pp. 167-175, 1914; 
xix, pp. 23&-237, 1914. 

8.^ 



84 Mycodextran Content of Aspergillus niger 



washed several times with alcohol, and finally dried to constant 
weight at 100°. The product was a hard white crust which was 
easily removed from the paper. The results obtained as alx)ve 
are recorded in the following table. 



1 

1 
AGE or CULTURE 


WEIGHT OP DRY 
FUN0U8 


MYCODEXTRAN 


(DkY FUNGUS) 


dys. 


gm. 


gm. 


prr cen 


3 


13.648 


Trace 




5- 


20.096 


0.177 


0.88 


7 


13.977 


0,666 


4.76 


14 


8.026 


0.589 


7.34 


21 


7.537 


0.570 


7.56 


28 


6.915 


0.505 


7.30 


35 


7.002 


0.549 


7.84 


42 


6.734 


0.546 


8.11 


49 


6.441 


0.541 


8.45 


56 


6.171 


0.523 


8.47 


63 


6.331 


0.544 


8.59 


142 


5.938 


0.546 9.19 

1 



In the three day cultures a dense white mycelium had com- 
pletely covered the surface of the medium, but spores had not 
yet begun to form. The five day cultures showed quite a few 
black spores and a great number of yellow spores in wliich the 
pigment had not yet developed. The week old cultures were jet 
black on the surface. At this stage autolysis had already begun, 
as indicated by the decrease in weight of the dry fungus, and from 
this point on a gradual decrease in weight occurred, similar to that 
reported for Aspergillus fumigatus in another paper.- As autolysis 
progressed, the hot water extract became lighter in color, while, 
on the other hand, the medium became darker. Much to our 
surprise, the yield of mycodextran was found to be approximately 
constant for a given amount of (Culture medium. The percentage 
yield, of course, increased as the weight of the fungus diminished. 
Our previous failure to obtain mycodextran from cultures of Pern- 
clUium expansum that had autolyzed for six weeks, we arc at pres- 
ent unable to explain. In the light of this later work it appears 
probable that mycodextran forms an integral part of the mature 
fungus and does not undergo autolysis. 

' A. W. Dox: this Jonrnnl, xvi, p. 483, 1913-14. 



Arthur W. Dox 85 

There is apparently a relationship between the presence of 
spores and the occurrence of mycodextran. From the three day 
cultures, which were perfectly white, the merest trace of myco- 
dextran was obtained, whereas the five day cultures, in which 
spores were beginning to develop, showed an appreciable amount 
of this substance, and all the cultures that had grown seven days 
or longer and were black with spores gave the maximum yield. 
Since the spores retain their vitality for a long time, it is not Ukely 
that they undergo autolysis. The fairly constant yield of myco- 
dextran might, therefore, be explained by assuming that this 
substance occurs only in the spores. Cramer^ obtained a yield of 
17 per cent of "spore starch" from the spores of PenidUium glau- 
cum. This spore starch, we have reason to believe, was probably 
an impure preparation of mycodextran contaminated by some 
other carbohydrate giving the iodine reaction. When Aspergillus ^ 
niger is grown on a fluid medium, such as Raulin^s, the surface 
collects considerable moisture, and a separation of the spores from 
the hyphae is a difficult matter. Our evidence in favor of the 
assumption that mycodextran is locaUzed in the spores of this 
organism is, therefore, entirely presumptive, until we succeed in 
separating a sufficient quantity of spores with which to make an 
extraction. 

For preparing mycodextran in quantity, cultures that have un- 
dergone autoljrsis are preferable. Not only can the maximum 
jdeld be obtained from such cultures, but the product is more free 
from contaminating substances, and, therefore, more easily puri- 
fied. 

Whether or not mycodextran occurs as such in the fungus can- 
not be definitely stated at this time. The method of extraction, 
^iz,f boiling with water, might, of course, bring about a partial 
hydrolysis of some mother-substance, resulting in the formation 
of this peculiar polysaccharide. 

» E. Cramer: Arch. f. Hyg,, xx, p. 197, 1894. 



ON THE FAT IN THE BLOOD IN A CASE OF LIPAEMIA. 

By C. G. IMRIE. 
(From the Department of Pathological Chemistry of the University of Toronto.) 

(Received for publication, November 28, 1914.) 

A patient was admitted into the Toronto General Hospital in 
March, 1914, who up to that day had been at work as a labourer, 
and eight hours after admission died in a state of coma. 

At the autopsy it was observed that the blood that escaped in 
the first incision was remarkably pale in colour, appearing like 
cream on dividing vessels in the least dependent parts of the 
cadaver; in more dependent parts it was redder, but still evidently 
rich in fat. In the pancreas a nodular projection was found in 
front of the spinal colunrm, and extensive induration, interlobular in 
distribution and of a more chronic character, stretched through- 
out the tail of the organ. A full account of the microscopical 
anatomy of this, as well as of a remarkable condition in the spleen 
of this patient, is given elsewhere by Prof. J. J. Mackenzie.^ Blood 
was collected from the heart and great vessels of the thorax, and 
urine from the bladder. 

In the urine sugar was found amounting to 6.2 per cent; and 
from 50 cc, acidified with sulphuric acid and saturated with am- 
monium sulphate, ether in a continuous extraction apparatus ex- 
tracted a laevorotary acid which in its rotatory power was equiva- 
lent to 0.67 per cent of jS-hydroxy butyric acid in the urine. The 
ammonia coefficient was 11.9. In addition, ether extracted from 
the urine 0.1 per cent of acids insoluble in water, the molecular 
weight of which by titration was found to be 282, and which in 
other way^ too appeared to be higher fatty acids. With them was 
obtained also a trace of cholesterol amounting to about 3 mgm. 
from 50 cc. of urine. 

The blood was put in a separating funnel and the creamy serum 
separated from the corpuscles. Its specific gravity at 16.5°C. 

* J. J. Mackenzie: unpublished report. 

S7 



88 Fat in the Blood in a Case of Lipaemia 

was 1014. Of this serum 109.5 grains were treated with 5 cc. of 
40 per cent potash and about 100 cc. of petroleum ether and as 
much alcohol. After being shaken, the clear petroleum solution 
was removed and the shaking was repeated three times with fresh 
portions of petroleum ether. A fifth portion of petroleum ether 
used in the same way was found to have removed only 2.5 mgm. of 
soluble material. The mixed petroleum solutions were made up 
to 250 cc. In a portion of this the dissolved matter was deter- 
mined to be 14.06 per cent of the serum by weight. This residue 
was saponified with alcoholic potash and the alkaline solution shak- 
en with petroleum ether several times. Unsaponifiable substances 
thus removed amounted to 1 .5 per cent of the serum by weight. 

The alkaUne serum was then made acid and again shaken with 
petroleum ether, which now removed a further quantity of soluble 
matter, fatty acids, originally present in the serum as soaps, and 
Amounting to 0.38 per cent of the serum. 

Thus there were found in the serum: 

A. Neutral fat and cholesterol, 14.06 per cent. 

B. Cholesterol (unsaponifiable matter), 1.5 per cent. 

C. Fatty acids originally in the form of soaps, 0.38 per cent. 
Fraction A, Neutral fat and cholesterol gave the iodine value, 

by Wijs's method, 78.6, and the saponification value, after deduct- 
ing cholesterol, 190; the amount of insoluble fatty acids obtained 
by saponification of this portion was 91.2 per cent of the neutral 
fat, and these acids then had the iodine value 73.0. The products 
of saponification soluble in acid obtained from 2 grams of the ex- 
tract were boiled with nitric acid for some time and examined for 
phosphoric acid, but no phosphomolj'^bdateof ammonia was fonned. 
Fraction By containing the cholesterol, was remarkably colour- 
less and crystallized as if it were almost pure cholesterol. A solu- 
tion of it in alcohol containing 0.2237 of a gram was precipitated 
with digitonin by Windaus's method, and yielded a precipitate of 
the cholesteride corresponding to 0.2190 of a gram of free choles- 
terol; so esters of cholesterol cannot have been present in appre- 
ciable amounts. The iodine value of the unsaponifiable matter 
when first obtained was 118 by Wijs's method; seven months 
later when determined again by this method it had fallen to 89, 
but by Hvibrs method the theoretical value for pure cholesterol 
was then given; t;/z., 65. The melting pomt of the unsaponifiable 



C. G. Imrie 89 

matter without being purified by recrystallization was 141-142°C. 
Pure cholesterol melts at 145-146°. 

Fraction C, the fatty acids originally present in the serum as 
soaps, gave the iodine value 88.6. 

Some of the subjacent corpuscular layer was heated with potash 
and alcohol, as in liebermann's method of fat estimation; on 
acidification and shaking with petroleum, fatty acids were obtained 
amounting to 0.15 per cent, which is not more than is found in 
normal blood corpuscles. 

1 . The fat in the blood of this case was present, therefore, in 
large amount: larger amounts have been recorded; for instance, 
18 per cent (Fischer),^ 19.7 per cent (Neisser and Derlin),^ 29 per 
cent (Adler).^ But it was evident that the amount could vary 
considerably according to the level in the body of the vessels 
from which the blood was obtained. 

2. The fat was composed almost entirely of simple glycerides 
and contained but Uttle if any of the phospholipines, such as lecithin. 
For on saponification the fat yielded 91.2 per cent of insoluble 
higher fatty acids, whereas lecithin with the conventional formula 
yields at most 70 per cent, other phospholipines less still, and the 
simple glycerides of adipose tissue about 95 per cent. The fact 
that less than this amount was obtained may be explained by sup- 
posing that a small amount of some lower soluble acids, such as 
butyric acid, took part in the composition of the glycerides, and 
need not indicate the presence of phospholipines, the absence of 
which was made clear by the failure to detect any phosphoric acid. 

3 . The fatty acids entering into the composition of the fat, as 
shown by their iodine value 73, were similar to those obtained 
from adipose tissue (iodine value about 65) and quite different 
from those normally found in the liver, heart, or other organs. 
The liver of this same patient gave 5.7 per cent of fatty acids, by 
the same treatment and method of estimation, ha\ing the iodine 
value 104 (this fat in the liver was much more conspicuous in the 
capillaries than in the cells), the heart gave 2.36 per cent, with 
iodine value 132, and the kidney 3.72 per cent, a large amount for 
this organ, with iodine value 105. These data point to the cause 

* B. Fischer: Virchows Arch. f. path. Anat., clxxii, pp. 30 and 218, 1903. 

* E. Neisser and L. Derlin: Ztschr. f. klin. Med., li, p. 428, 1904. 
« M. Adlcr: Berl. klin. Wchnachr., xlvi, p. 1453, 1909. 



90 Fat in the Blood in a Case of Lipaemia 

of the lipaemia in this case being the mobilization of connective 
tissue fat; the other possibility, accumulation in the blood of fat 
absorbed from the food, is rendered improbable by the large amount, 
which in the whole blood can hardly have been less than 300 grams 
and was probablj' more. 

4. The amount of soap in the blood is striking; though in com- 
parison with the amount of neutral fat it is small, absolutely as 
much fatty acid was present in this form as is often present in all 
kinds of combination in normal blood. 

5. The amount of cholesterol, 1.5 per cent of the serum, rather 
more than one part for every ten parts of fat in the blood, is also 
striking; but still larger amounts have been found; e.g., 2.6 and 
3.6 per cent (Adler). 

6. No evidence of the presence of lecithin in the serum was ob- 
tained. The tendency for the amount of lecithin in the blood in 
diabetic Upaemia to be small has been pointed out by Adler and 
by Klemperer and Umber .^ 

The above work was carried out under the direction of Prof. 
J.«B. Leathes, to whom I am much indebted for his kind interest 
and suggestions. 

* G. Klemperer and 11. Umber: Ztnchr. f. klin. Med., Ixv. j). 340, 1908. 




FIBRIN. 

By a. W. B08W0RTH. 

{From the Boston Floating Hospital^ Boston, and the Chemical Laboratory oj 
the New York Agricultural Experiment Station, Geneva.) 

(Received for publication, November 30, 1914.) 

Certain observations made while working with blood seemed 
to indicate that fibrin might possess some chemical properties 
quite similar to those of casein, which have been reported by 
Van Slyke and Bosworth.^ In order to investigate this question 
it lx?came necessary to prepare pure ash-free fibrin. 

Method of preparing a^h -free fibrin. Fresh ox blood was collected in a large 
bottle and carried immediately to the laboratory where it was transferred to 
precipitating jars and allowed to coagulate. The clot« were removed, broken 
into small pieces, and washed in running water to remove the serum and 
blood corpuscles. The washed masses of fibrin were passed through a meat 
chopper, placed in an 8 liter bottle, a little toluene was added, and the bot- 
tle filled with a 0.2 per cent solution of sodic hydrate. This solution caused 
the fibrin to swell, and after about 36 hours the whole contents of the bottle 
resembled a thin jelly. This jelly was broken up, one-half transferred to 
another 8 liter bottle, and after the two bottles were filled by the addition 
of water they were allowed to stand another 36 hours. The jelly was al- 
most completely dissolved, so the contents of the two bottles were filtered, 
first through cheese cloth, then linen, and finally paper. The clear filtrate 
was divided into several portions which were placed in 2 liter precipitating 
jars, diluted with an equal volume of water, and the fibrin was precipitated 
by the cautious addition of 0.3 per cent acetic acid. At a certain point a 
fiocculent precipitate appeared which quickly settled to the bottom of the 
jar. 

The suiMjrnatant liquid was poured off, the precipitate washetl with water, 
dissolved in dilute sodic hydrate (0.05 per cent), and again precipitated 
with acetic acid. This process was repeated three times, the final precipi- 
tate being washed with alcohol and ether and dried over sulphuric acid in 
an evacuated desiccator. The fibrin strongly resembled casein in all stages 
of its preparation except in its extreme sensitiveness to a slight excess of 
acid or alkali, for unlike casein it is readily soluble in weak acetic acid. 



» This Journal, xiv, pp. 203-206, 1913. 

91 



92 Fibrin 

The final product was a very fine, light, white powder which gave the 
following figures upon analysis. 

per cent 

Moisture 1.S4 

Ash in dry substance .03 

Carbon in dry substance .51 .82 

Hydrogen in dry substance 6 .90 

Nitrogen in dry substance 17 .21 

Sulphur in dry substance 0.05 

Oxygen in dry substance* 23 . 12 

•By difference. 

In order to show that the preparation obtained was not a sub- 
stance or mixture of substances i-esulting from the hydrolysis of 
fibrin by the alkali used to dissolve the fibrin, the following experi- 
ment was performed. 

Some of the fibrin, prepared as above, was dissolved in 0.2 per cent sodic 
hydrate, allowed to stand for 24 hours at 37°C., and the fibrin precipitated 
by means of acetic acid. The x)recipitate, after being washed with water, 
alcohol, and ether, was found to give the same figures upon analysis as the 
original preparation. The filtrate contained soluble nitrogen which could 
not be precipitated by acids, showing that hydrolysis had occurred, but 
that the products of the hydrolysis did not contaminate the final preparation 
of fibrin. 

THE RELATION OF FIBRIN TO BASES.- 

Combinatiom with sodium. One gram of fibrin was found to 
require 30.7 cc. of yiy sodium hydroxide to make it neutral to 
phenolphthalein, or one gram of fibrin combines with 6.14 X 10~^ 
gram equivalents of sodium to form a compound neutral to phenol- 
phthalein. 

One gram of fibrin dissolved in 50 cc. of -^js sodium hydroxide 
requires the addition of 42.4 cc. of /o HCl to produce the first 
sign of a precipitate, or one gram of fibrin combines with 
1.52 X 10^ gram equivalents of sodium to form a compound 
soluble in water. 

One gram of fibrin dissolved in 50 cc. of t^V sodium hydroxide 
i*equires the addition of 50 cc. of -^ HCl to cause the complete 
precipitation of the fibrin. This proves that fibrin does not form 
an insoluble salt containing sodium. 

* For the details of the technique involved in these studies consult the 
work of Van Slyke and Bosworth upon the caseinatcs (loc. cii.). 



A. W. Bosworth 93 

Cmnbinaiions with calcium. One gram of fibrin was found to 
require 30.8 ec. of -^^ calcium hydroxide to make it neutral to 
phenolphthalein, or 1 gram of fibrin combines with 6.16 X 10~^ 
gra m equivalents of calcium to form a compound neutral to phenol- 
phthalein. 

One gram of fibrin dissolved in 100 cc. of /ly calcium hydroxide 
requires the addition of 77.5 cc. of -^ HCl to produce the first 
sign of a precipitate, or one gram of fibrin combines with 
4.50 X 10"^ gram equivalents of calcium to form a compound sol- 
uble in water. 

One gram of fibrin dissolved in 100 cc. of ^ calcium hydroxide 
requires the addition of 70.0 cc. of -^ HCl to cause a complete 
precipitation of the fibrin, or one gram of fibrin combines with 
3.0 X 10"^ gram equivalents of calcium to form a compound in- 
soluble in water. This precipitate is completely dissolved by 
5 per cent solution of sodium chloride.' 

Relation of fibrin to hydrochloric add. 

One gram of fibrin dissolved in 50 cc. of -^js HCl requires the 
addition of 42.5 cc. of ^ sodium hydroxide to produce the first 
sign of a precipitate, or one gram of fibrin combines with 
1.50 X 10"^ gram equivalents of hydrochloric acid to form a com- 
pound insoluble in water. 

Effect of carbon dioxide on solutions of fibrinates. 

Fibrin is not precipitated by running a stream of carbon dioxide 
into solutions of sodium, potassium, or ammonium fibrinates, 
while solutions of calcium or barium fibrinates give precipitates 
of acid fibrinates. Solutions of calcium fibrinate clot upon being 
exposed to the air, due to the absorption of carbon dioxide. 

Fibrin, unlike casein, does not decompose calcium carbonate 
when these two substances are triturated together in the presence 
of water. 

' Ca.sein forms a similar compound with calcium (this Journal, xiv 
p. 231, 1913). 



94 Fibrin 

Molecular weight of fibrin. 
From the sulphur content we have, 

32.07 
Mol. Wt. = X[ --rr X 100] » N 3375.7. If N= 2, then Mol. \Vt. - 6751 .4. 

U.UO 

From the sodium fibrinate containing one equivalent of base 
we have, 

1 
Mol. Wt. - ;-r;:7,7r. » 6666.6. 

1.0 X i\J~* 
CONCLUSIONS. 

1 . Pibrin can combine with both bases and acids to form definite 
compounds. 

2. Fibrin combines with four equivalents of base to form a 
compound which is neutral to phenolphthalein. 

3. Fibrin combines with bases to form a series of three acid 
salts which contain one, two, and three equivalents of base respec- 
tively. 

4. All the combinations of fibrin with sodium, pota^ium, and 
ammonium are soluble. 

5 . The calcium fibrinates containing three and four equivalents 
of calcium are soluble, the calcium fibrinates containing one and 
two equivalents of calcium being insoluble* 

6. Fibrin combined with one equivalent of acid is insoluble, 
and combined with more than one equivalent of acid is soluble. 

7 . Pure fibrin unlike casein is not strong enough as an acid to 
decompose calcium carbonate. 

8. The molecular weight of fibrin is about 6666. 

9 . Carbon dioxide precipitates fibrin from a solution of CpJcium 
fibrinate, but not from a solution of sodium, potassium, or am- 
monium fibrinate. 




Richter Rotary Extraction Apparatus 

A modlflcAlion of Iht »ppzr*tas deMoped In the Uborttoty of 9Vo/. 
GtalMlf Emdtn-Frtnkfort, •whtrt U is ased luUh mack saectss 

The apparatus is especially desirable for use with liquids from which the 
solids cted with difficulty, for instance, oxy-butyric acid can be ex- 

tnctet s, instead of 72 hours— the time required vith other apparatus. 

Appan ete with eloctric motor (variable speed), 3 heat 

, plate, mounted on support as illustrated . net $IS.M 

two extractora mounted on one support, complete, net $K.M 

Extrai -Btua — glass parts only net $ 7.50 

I lER & Amend 

FOUNDED 1851 
H I [or Educalioiul and Industrial Labontory Suppliu 




PITTSBURGH, PA. 

3006 J«nkint-Arcadc BIdg. 



HAMBVRQ, QBRMANY 



(CONTENTS 



\V. B. Thompson: ^>tu()ie8 in the blood rclatioimhiu of animals as displayed 
in tlie composition of tln^ soium proteins. III. A comparison of the 
sera of the hen, turkey, duck, and goose with respect to their content 
of various proteins 1 

R. S. Brioos: Studies in the blood relationship of animals as displayed in 
the composition of the serum proteins. IV. A comparison of the sera 
of the pigeon, rooster, and guinea fowl with respect to their content 
of various proteins in the normal and in the fasting condition 7 

Paul J. IIanzlik: Precipitation of serum-albumin and glutin by alkaloidal 

reagents 13 

Walter Jones and A. E. Richards: Simpler nucleotides from yeast nucleic 

acid 25 

Sergius Morgulis : Studies on fasting flounders 87 

Joseph C. Bock and Stanley R. Benedict: An examination of the Folin- 

Farmcr method for the colorimetric estimation of nitrogen 47 

Robert C. Lewis anri Stanley R. Benedict: A method for the estimation 

of sugar in small quantities of blood 61 

F. B. Kingsbury and E. T. Hell: 1 he svnthesis of hippuric acid in experi- 
mental tartrate nephritis in the rabbit 78 

Arthur W. Dox : The soluble polysaccharides of lower fungi. III. The in- 
fluence of autolysis on the mycodextran content of Aspergillus niger,, 88 

C. G. Imrie : On the fat in the blood in a case of lipaemia 87 

A. W. Bo8wor*TH : Fibrin 91 

P. A. Levene and F. B. LaForge : Correction. On chondroitin sulphuric 

acid. Second paper 05 



The Journal op Biological Chemistry is designed for the prompt publica- 
tion of original investigations of a chemical nature in the biological sciences. 
It is issued monthly. Each volume consists of as many numbers ai are required 
to make a total of between five and six hundred pages. 

The price of The Journal to subscribers in the United States and Canada 
is $3.00 per volume: in other countries. $3.25. Remittances should be made by 
draft or chock on New York or by postal money order, payable to Tbv Journal 
or Biolootcal Chemistry, and should be sent to 2410--21 Greenmount Avenue, 
Baltimore, Md., or to The Rockefeller Institute for Medical Research, 66th 
Street and Avenue A. New York City. 

Pr. Dak in is at pre^?ent abroad. Manuscripts may be sent to any other of 
the editors. Contrihutori:! will receive 100 copies of their papers free of charge, 
and additional copies will be furnished at cost. It is understood that, unless 
expressly agreed to the contrary, all manuscripts submitted to Tbb Joubnal oy 
Biological Chemistrt shall not be offered for prior or simultaneous publica- 
tion elsewhere. 

The price of this number is $1.00. 



THE WAVERLY PRESS 

■ ALTiMOfir. U. •. A. 



J 



VOL. XX FEBRUARY, 1915 No. 2 



THE JOURNAL ':^^--^ 

FEB I 7 



OF 



UfS. Oepafj/|.e«,{ 



BIOLOGICAL CHEMISTRY 

rOUITDXD BT CHKICTIAlf A. BEKTCK AND SUtTArNED IN PART BT TBB CBBMTIAN A. RBBTBB 

MB If OBI AL rUMO 

EDITED BY 

H. D. DAKIN, New York City. LAFAYETTE B. MENDEL, New Haven, Conn. 
E. K. DUNHAM, New York City. A. N. RICHARDS, PhUadelphU, Pa. 

DONALD D. VAN SLYELE, New York City. 

WITH THE COLLABORATION OF 

J. J. ABEL. BiltlmofB, Md. J. B. LSATHBS. Tortnto, Cauda. 

R. H. CHUTBNDBN, New Htftn, Conn. P. A. LBVBNB, New York. 

OTTO FOUN» BoBtoa, IUm. JACQUBS LOBB, New York. 

WILLIAM J. GIBS. Now York. A. S. LOBVBNHART, MadiMn, Wia. 

L. J. HBNDBRSON, CamMdge. Kbm. GRAHAM LUSK, Now York. 

RBID HUNT, Boiton, Man. A. B. MACALLUM, Toronto, Canada. 

W. A. JACOBS, New York. J. J. R MACLEOD. Clofolaod, Ohio. 

WALTBR JONBS, Baltimore, Md. JOHN A MANOEL, Now York. 

J. H. KASTLB, Loxinffton, Ky. A. P. MATHBWS. Chicago, 01. 

F. O. NOVY, Aao Arbor, Mich. 

THOMAS B. OSBORNB. New Hafoo, Conn. 

T. BRAILSPORD ROBBRTSON, Borkoloj, Cal. 

P. A. SHAFFER. St. Louli. Mo. 

A. B. TAYLOR. Philadelphia. Pa. 

F. P. UNDBRHILL, New Hafon. Conn. 

V. C. VAUOHAN, Ann Arbor, Mich. 

ALFRED J. WAKEMAN. New Haren. Conn. 



PUBUSHED MONTHLY 
BY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH 

FOR THE JOURNAL OF BIOLOGICAL CHEMISTRY, INC. 

2419-21 GREENMOUNT AVENUE. BALTIMORE. MD. 

Entered ao i»econd-claaa matter, August 1. 1011. at the Post OrTice at Baltimore. Md.. under ilie 

Act of Mareh 8. 1870. 



COPTRIQHT 1015 
BY 

Thx Journal or Biological Chemibtbt 



\ 



AN IMPROVED METHOD FOR THE ESTIMATION OF 

INORGANIC PHOSPHORIC ACID IN CERTAIN 

TISSUES AND FOOD PRODUCTS.^ 

By ROBERT M. CHAPIN and WILMER C. POWICK. 

{Prom the Biochemic Division, Bureau of Animal Industry, U, S, Department 

of Agriculture, Washington.) 

(Received for publication, December 12, 1914.) 

INTRODUCTION. 

For the solution of the much discussed questions as to the pre- 
cise significance of the organically and of the inorganically com- 
bined phosphorus in animal and in vegetable tissues, a method for 
sharply and surely differentiating between these two classes of 
compounds and for estimating accurately the amount of phos- 
phorus combined in each is obviously essential. With a view to 
supplying this need, many methods, differing among themselves 
in accuracy and in practicabiUty, and each more or less specifi- 
cally adapted to the investigation of a certain limited range of 
materials, have been evolved; but as yet no method has been found, 
if indeed one ever can be found, that is universally superior in its 
application to all substances. The improved method here offered, 
however, — designed for and successfully used in the determination 
of inorganic phosphorus in eggs and in meats, — is believed to possess 
a sufficient number of points of superiority over its immediate 
competitors, together with a suflSciently wide range of applica- 
bility, to justify its pubhcation at this time. 

DESCRIPTIVE PART. 

Discussion of improvements dmmed for the method. 

Of the methods now available for the determination of inorganic 
phosphorus in flesh, there are three which seem to stand pre- 

* Published by permission of the Secretary of Agriculture. 

97 



jouMKAL ormoLoaMCAL CHEMtamr, vol, XX, no. 2 



S 



98 Estimation of Inorganic Phosphoric Acid 

eminent; viz.j the Emniett and Grindley method,* the Siegfried 
and Singewald method,' and the Forbes method* — each of which 
has undoubtedly proved reasonably satisfactory in its particular 
field. Yet their failure to guard sufficiently well against tedious 
filtrations, decomposition of organic phosphorus compounds, and 
incomplete separation of the organic from the inorganic phos- 
phorus, makes problematical the accuracy and practicability of 
each when applied to tissues other than flesh, or even to that 
tissue after the occurrence of degenerative changes. 

The requirements of a generally trustworthy method for the 
determination of inorganic phosphorus in physiological tissues, the 
shortcomings of the above mentioned methods, and the advan- 
tages claimed for the improved method are briefly summarized 
below. 

Extraction. To insure the solution of any water-insoluble phos- 
phates that might be present, the extracting medium should be 
acid. To avoid any possible chemical decomposition of the or- 
ganically combined phosphorus, the use of strong reagents and of 
heat should be avoided. As a precaution against bacterial and 
enzjrmatic changes, the extraction should be effected quickly, in 
the cold, and in the presence of an antiseptic. Most simply to 
prevent the interference of colloids in the subsequent phosphate 
precipitation and the unnecessary exposure of the organic phos- 
phorus compounds to the action of the precipitant, these sub- 
stances should be rendered initially insoluble in the extracting 
medium, either by physical or by chemical methods. For obvious 
reasons, time-consuming operations should be avoided as far as 
possible. 

/' The Emmett and Grindley method falls short of the above re- 
Piuirements by the use of a neutral solvent and of heat; the Sieg- 
fried and Singewald method, by failure to remove the protective 
colloids, by exposure of the organic phosphorus to the action of 
the phosphate precipitant, and by the long duration of the initial 

' A. D. Emmett and H. S. Grindley: Jour, Am. Chem. Soc, xxviii, p. 25, 
1906; H. S. Grindley and E. L. Ross: this Journal, viii, p. 483, 1910. 

* M. Siegfried and E. Singewald : Ztschr. f. Untersuch. d, Nahrungs u, 
Genussvdttelj x, p. 521, 1905. 

«E. B. Forbes, A. Lehmann, R. C. CoUison, and A. C. Whittier: Bull. 
215, Ohio Agric, Exper. Sta,, 1910. 




R. M. Chapin and W. C. Powick 99 

filtration; and the Forbes method, by the use of heat and the neces- 
sity for a double filtration. Finally, Collison's* modification of 
the Forbes method of extraction is open to the criticism that the 
proteid superficially coagulated by the strong alcohol interferes 
with the further penetration of the tissue by the solvent. 

In the method herein proposed, however, the above require- 
ments have been met by employing as the extracting medium an 
aqueous solution of picric acid containing a small amount of hy- 
drochloric acid./ The extraction is complete; bacterial action is ; 
prevented and the proteids are effectually coagulated by the re- \ 
agent; whUe the danger of chemical or enzymatic changes is mini- \ 
mized by the low temperature and the rapidity of the extraction. / 
The extract is easily filtered and is practically free from organically 
combined phosphorus; while by the use of an aliquot of the fil- , 
trate for further work, a tedious washing of the precipitate is 
avoidedy^ 

Separation of the organic from the inorganic phosphorus. In 
view of the meagemess of our knowledge regarding the organic 
phosphorus compounds in physiological tissues, the separation of 
these from the inorganic phosphates must be conducted largely 
according to a priori considerations. Upon such grounds it would 
appear that the best assurance of a quantitative separation of the 
inorganic from the organic phosphorus and from other contami- 
nating substances is to be found in an at least partial separation 
at the time of the extraction (which we have secured by the use 
of picric acid) and a subsequent double precipitation of the inor- 
ganic phosphates, once from aeid and once from alkaline solution. 
Obviously, the precipitant first to be employed should be that 
which is least likely to decompose the organic phosphorus com- 
p)ounds,or,by precipitating them, to expose them to decomposition 
in subsequent operations; while the final precipitant should be that 
which yields the precipitate best adapted for the estimation of 
small amounts of phosphoric acid. 

Since, in the initial use of ammonium molybdate, — ^the only 
available acid precipitant,— the presence of strong reagents and 
the precipitation of more or less of the organic matter are in- 
volved, and since in its use as a final precipitant a compara- 
tively heavy precipitate is obtained from a relatively small amount 

» R. C. CoUison: Jour, Ind. and Eng. Chem., iv, p. 606, 1912. 



/ 



l^ 



lOO Estimation of Inorganic Phosphoric Acid 

of phosphoric acid, the initial use of an alkaline precipitant is 
clearly indicated. Of these, magnesia mixture seems to be pref- 
erable to either barium chloride or calcium chloride, for by its 
use a clear filtrate is easily obtained and the use of sulphuric acid 
in subsequent operations is not rendered impossible. 

In the method here proposed, therefore, the initial precipitation 
has been made by means of magnesia mixture and the final precipi- 
tation by means of ammonium molybdate, which, we believe, 
insures for the method a further superiority over the Emmett 
and Grindley method, in which the order is reversed. 

Final estimation of the inorganic phosphoric acid. Unques- 
tionably the inorganic phosphorus should be determined directly 
and not obtained by difference; for it is the form of phosphorus 
concerning which the most precise information is usually desired, 
and it is Ukewise the form which is most amenable to exact deter- 
mination. 

There has been much study, resulting in a voluminous literature, 
upon the final estimation of phosphoric acid; but as a criticism or 
comparison of the methods from time to time proposed does not 
lie within the scope of this paper, the reader is referred to the 
work of Artmann,® which contains a critical review of the methods 
now available. The present writers have chosen to make their 
final estimation gravimetrically, by the direct weighing of the 
ammonium phosphomolybdate: first, because the previous pro- 
cedure in their method leads most logically (as has already been 
pointed out) to this method of final estimation; and secondly, 
which is more important, because the precipitate of ammonium 
phosphomolybdate is much larger (nineteen times heavier) than 
a chemically equivalent precipitate of magnesium pyrophosphate, 
an important consideration in view of the small amount of phos- 
phoric acid usually dealt with. 

The particular method of estimation that we have chosen is 
that of Lorcnz,' which has been used with satisfaction by a num- 
ber of workers, including especially Neubauer and Liicker,® who 
have proved its reUability and have suggested slight modifications 
in its technique in the interests of convenience and economy. 

• P. Artmann: Ztschr. Aitgew, Chcm.y xxvi, pt. I, p. 203, 1913. 

^ N. V. Lorenz: Die landwirischaft. Versuchsst.f Iv, p. 183, 1901; Osierr. 
chem, Ztg.f xiv, p. 1, 1911. 

• H. Neubauer and F. Liicker: Ztschr,/. anal. Cheyn.y li, p. 161, 1912. 



R. M. Chapin and W. C. Powick loi 

Detailed execution of the method. 

Extraction. Rejecting the method of extracting the phosphoric 
acid by exhausting the tissue by repeated washings, we have chosen, 
in the interests of convenience and simpUcity and in order to se- 
cure a more concentrated extract, to subject the material to a 
single, thorough maceration with the solvent, and to employ an 
aliquot M the filtrate for further work. We have devised three 
methods by which to determine the exact concentration of the 
extract in terms of the original material, thereby avoiding errors 
due to the presence of an originally unknown volume of water 
or of insoluble matter in the sample. These methods are de- 
scribed below as Modifications A, B, and C. 

Modification A. This modification is applicable when the 
water content of the sample is known or can be determined with 
sufficient accuracy. 

A weighed sample of the material, containing between 8 and 80 
mgm. of inorganic P2OB, is macerated in a porcelain mortar with 
20 grams of dry, acid-washed sand, and is thoroughly, but quickly, 
mixed with an accurately measured amount of water (about 
200 cc.) and 10 cc. of 2.5 normal hydrochloric acid. The mixture 
is transferred to a wide mouthed, glass-stoppered bottle of about 
500 cc. capacity, and 5 to 8 grams of powdered picric acid are 
added. The bottle is shaken at frequent intervals during the 
next two hours, or continuously, by machine, for one-half hour; 
after which the extract is filtered through a folded filter, and 
100 cc. of the filtrate are measured out for subsequent work. 

The volume of the whole solution is found by adding to the 
volume of water and hydrochloric acid employed the volume of 
water contained in the sample; and the proportional amount of 
the original sample represented by 100 cc. of the filtrate is then 
calculated. 

Modification B. This modification is applicable when the 
volume of the insoluble matter is known or is negUgiblc. 

The weighed material is ground with sand in a mortar, as under 
Modification A. The whole is then diluted and quantitatively 
transferred with water to a graduated flask or stoppered cylinder, 
where it is treated with 10 cc. of 2.5 normal hydrochloric acid 
and 5 to 8 grams of powdered picric acid as before. It is then 



I02 Estimation of Inorganic Phosphoric Acid 

diluted with water to a definite total volume, and shaken, filtered, 
and sampled, as described under Modification A. 

To determine the volume displaced by the insoluble matter, 
the residue is sucked dry, its weight and specific gravity are deter- 
mined, and its volume is calculated. If the use of sand is un- 
necessary and the amount of insoluble matter is negligible, 
obviously the volume correction may be omitted; while if sand is 
employed, but the volume of the insoluble matter is negUgible, 
only a once for all determined correction for the volume of the sand 
need be applied, provided, however, that in each case the picric 
acid is added after the apparent volume of the solution has been 
observed. 

Modification C. In this case a "marker" is used to indicate 
the concentration of the resultant extract, on the principle that 
if a definite amount of some inert chemical that is not naturally 
present in the material under examination be added at the begin- 
ning, a determination of the concentration of such chemical in 
the final extract will afford a measure of the total volume of 
liquid in which this chemical, and therefore the PsOs, is dissolved. 
A half normal solution of potassium iodide has been chosen for 
this puri)ose, and the urea-nitrite method of Schirmer* has been 
used for the determination of the potassiimi iodide in the filtered 
extract and for the standardization of the half normal potassium 
iodide solution. The method of Kendall,*® which was first tried, 
proved to be inapplicable in the presence of picric acid, while 
the other methods of Schirmer were less suited to our purpose 
than the method chosen. 

In this modification the weighed material is ground with sand 
as under Modification A, and by the aid of water is quantitatively 
washed into a 500 cc. glass-stoppered bottle and brought to a 
volume of about 200 cc. The mixture is then treated successively 
with 10 cc. of 2.5 normal hydrochloric acid, exactly 25 cc. of the 
half normal potassium iodide solution, and from 5 to 8 grams of 
powdered picric acid. The bottle is shaken and its contents 
filtered, as described under Modification A; and from the filtrate 
two portions of 50 cc. and 100 cc. are measured for the potassium 
iodide and for the phosphoric acid determinations, respectively. 

• W. Schirmer: Arch. d. Pharm., ccl, p. 448, 1912. 
*o E. C. Kendall: Jour. Am. Chem. Soc, xxxiv, p. 894, 1912. 



R. M. Chapin and W. C. Powick 103 

The 50 cc. portion is measured into a spacious, narrow necked, 
glass-stoppered bottle, treated with 10 cc. of five normal sulphuric 
acid and 15 cc. of a freshly prepared 1 per cent sodium nitrite 
solution, and after one to two minutes with 1.5 grams of crystal- 
lized urea. The solution is shaken vigorously in the tightly stop- 
pered bottle until the excess of nitrous acid is destroyed, after 
which the separated iodine is dissolved by the addition of 10 
cc. of normal potassium iodide solution and titrated in the usual 
manner against thiosulphate in the presence of starch. The 
thiosulphate equivalent of 5 cc. of the original half normal po- 
tassium iodide solution is determined in the same manner, and 
multipUed by five to obtain the thiosulphate equivalent of the 
potassium iodide originally added to the sample. If this figure 
be represented by A cc, and the thiosulphate required for the 
50 cc. aliquot of the filtrate by B cc, and the weight of the tissue 
operated upon by W gm., then the 100 cc. of the filtrate to be used 
for the PiOs determination will obviously represent ^^^ grams of 
the original material. 

Separatum of the inorganic phosphoric add. The phosphoric 
acid in the 100 cc aliquot of the filtrate obtained by either of the 
above modifications of the extraction process is precipitated in 
the usual manner with magnesia mixture and ammonia; and the 
precipitation is completed, or allowed to complete itself, by agi- 
tation or by sufficient standing. The precipitate is filtered off 
and washed with 2.5 per cent ammonia water until the washings 
are practically colorless; then the precipitate on the filter and 
in the beaker is dissolved in dilute nitric acid, and washed with 
water into a 150 cc beaker. The combined solution and wash- 
ings are then evaporated to dryness and the residue is redissolved 
in a mixture of 25 cc. of nitric acid (sp. gr., 1.20), 1 cc of con- 
centrated sulphuric acid, and 25 cc of water." 

Final estimation of the inorganic phosphoric acid. The final 
estimation of the inorganic phosphoric acid in the nitric-sul- 
phuric acid solution of the dry residue, prepared as above, we 
have carried out by the Lorenz method, with exact adherence 
to the directions which he has given, both for the precipitation 

^^ Ammoniacal alcohol might, of course, be used for the washing of the 
magnesium ammonium phosphate precipitate and acidified alcohol for dis- 
solving it, according to the method of Forbes; but this probedure has not 
appeared to be necessary in our work thus far. 



104 Estimation of Inorganic Phosphoric Acid 

and for the filtration and washing, except that in some cases we 
have used acetone for the washing, as suggested by Neubauer 
and Liicker, in place of alcohol and ether, as directed by Lorenz. 
The use of acetone for this purpose we have found to be perfectly 
satisfactory, provided only that the acetone is free from aldehydes 
and is otherwise pure. 

Discussion and explanation of the various steps of this method 
are to be found in the original references, and for the sake of 
brevity will be omitted here. 

EXPERIMENTAL PART. 

Experimental results in the development and application of the 

improved method. 

Lorenz method. Chemically pure sodium biphosphate was dis- 
solved in water, and the solution was standardized by weighing 
both the residue obtained by the evaporation of 100 cc. in a 
platinum dish, and the residue of sodium metaphosphate subse- 
quently obtained on ignition. Duplicate determinations all led 
to identical values for the concentration of the solution. 

The P2O6 factor of the precipitate yielded by the Lorenz method 
was determined by weighing the precipitates obtained from 
varying quantities of the above solution, conditions in duplicate 
determinations being purposely varied between the limits speci- 
fied by Lorenz without appreciable irregularities resulting. The 
method was found to be thoroughly satisfactory, and the P2O6 
factors determined by us differed from those given by Lorenz 
by no more than the experimental error, as shown in the following 
table : 

TABLE 1. 
PiOi factors for t^e vriih the Lorenz method. 



APPROXIMATE 


FACTOR 


WEIGHT or 






PRECIPITATE 


Found by ua 


Given by Loreni 


gm. 






1.5 


0.03307 


0.03291 


1.0 ; 


0.03305 


0.03295 


0.8 


0.03298 


0.03299 


0.4 


0.03312 




0.3 1 




0.03301 


0.12 


0.03322 


0.03307 



R. M. Chapin and W. C. Powick 



105 



Originally the factors obtained by us for small amounts of 
PsOs were appreciably lower than those shown in the table, but 
this was found to be due to a separation of MoOj that had taken 
place in consequence of an insufficient amount of nitric acid 
in the reagent. In such work as was done with this reagent, 
however, the factors experimentally found with that solution 
were used. Lorenz has recommended the general use of the factor 
0.03295, which, being based upon more exhaustive work than that 
done by us, we are inclined to accept in preference to our sUghtly 
different figures, particularly as the difference may be due to our 
having used a less pure salt. 

TABLE 2. 
Recovery of PtOifrom picric acid solution. 




P,0» used, gm 0.004106 

Weight of precipitate, gm 0. 1257 

Factor 0.03259 

PjO» recovered, gm j 0.004097 

PaO» recovered, per cent |99.77 



PjOs recovered, average per cent. 



0.004106 I 0.00410C: 0.006159 



0.1265 
0.03259 
0.004113 
100.11 



0.1257 I 0.1888 
0.03259 I 0.03259 
0.004097, 0.006153 
99.77 ,99.90 



99.89 



Effect of picric acid on the recovery of P^O^. Measured portions 
of the standard phosphate solution were diluted to 100 cc. with a 
saturated solution of picric acid, and precipitated with magnesia 
mixture and ammonia. After twelve hours the precipitates were 
filtered off and washed, dissolved in dilute nitric acid, and the 
phosphoric acid was determined by the method already indicated. 
The results are given in Table 2. 

Considering the small volume of the standard phosphate solu- 
tion that was used, the results shown in Table 2 are extremely 
satisfactory. 

Recovery of P^Oi added to eggs and to meat. In the next set of 
experiments a considerable quantity of whole egg substance was 
thoroughly mixed, and its water content was determined by 
drying a sample in a vacuum oven at about 65°C. Inorganic 
phosphoric acid determinations were then made upon portions of 
the original egg substance and likewise upon portions of the egg 



io6 Estimation of Inorganic Phosphoric Acid 

substance to which known amounts of the standard phosphate 
solution had been added, with a view to ascertaining how much 
of the added phosphoric acid could be recovered. The determi- 
nations were carried out by the method already described, the 
picric acid extract being made by Modification A. The results 
of this experiment are given in Table 3. 

In another similar experiment on meat (see Table 6) in which 
the extracts were made by Modification C, 98.73 per cent, 101.12 
per cent, and 98.97 per cent were recovered in triplicate deter- 
minations, making an average recovery of 99.61 per cent. 

The difference between the amounts of phosphoric acid added 
and the amount recovered in the above experiments is in each 
case small; and since the loss or gain is divided between two 
separate determinations, each of which, in the case of the experi- 
ment on meat, involves an iodine as well as a phosphoric acid 
determination, the errors appear to be well within the permissible 
limits. 

Experiments on the iodine method for dscertaining the concentra- 
tion of the extract. The first step in this series of experiments 
was to check the Schirmer method for the estimation of potassium 
iodide. 

A standard solution of potassium iodide was prepared from a 
weighed amount of the pure substance, and was checked by ti- 
tration against thiosulphate by Kendall's method. The standard 
solution of thiosulphate had been based upon standard dichromate 
in the usual way. It was calculated that 1 cc. of standard potas- 
sium iodide solution should be equivalent to 0.3288 cc. of the 
standard thiosulphate. The results of the tests on Schirmer's 
method, both in the presence and in the absence of picric acid, are 
given in Table 4. 

The Schirmer method having been shown to be satisfactorily 
accurate, the whole method of P2O6 determination according to 
Modification C was then carried out on a standard phosphate 
solution as follows: 

To 10 cc. of a standard phosphate solution, containing a total of 0.0154 
of a gram of PaOt, were added 25 cc. of a standard potassium iodide splution. 
To this mixture were added 10 cc. of a 2.5 normal hydrochloric acid solu- 
tion, 5 grams of picric acid, and sufficient water to make the volume about 
250 cc. Of the final solution two aliquot portions of 25 cc. were used for the 



R. M. Chapin and W. C. Powick 



107 



o 

O 

Q 
4 



n 



8 



8S 



;t S 



' 00 



n 



8 «^ S 



M 



:f CO o» tj 

CO -^ iQ O 

oQ Q m5 M 
ei 5 r« r« 

8858 



d 






1 s 



o . 

E 

a 

< 
o 



n 



8 



S^ 



?ff 



? I 8 



00 



P» r< ^ 



^ =^ o» - 

06 $ >0 M 
c« S r« t^ 

' ^8858 



S 



S I SS 



3 i 



000 



*=^s 



o t>. 



8 1 8 S 1^ 8 



8 



8 



?ff 



8S 



S 8S 



8 
8 



M 



CO 

00 



8 
S 



M 



:: 8 8 8 



Ok CO oo 

o d d d 



o 



9 






« 
9 



M 

9 



e to 



s 



8-» 



c< 



eo 



9^ s 



to 1^ 



M 



«4 o o 



O 00 

8S5 



•o 

s 



s 



8-^ 



in 



CO 

d 



s 



90 O t* 



M o 00 

80S 



eo 



CO 



s 



«0 






o 



a 
a 

< 

o 



8 ' 8 



8IO 
n I CO 

i?i I ^ 



M 



8 



8 



83 



eo 

d 



9' 



•o 



«s « t» 

Ok CO CO Qp 

<M kg ^ v^ e 

d d d d d 



9' 



•o 



Is! 



00 00 

CO o» 

8 



8 



8 



8S 
55 



CO 



n 
^ 



10 



^ a »o 

„ 0» CO ^ 

M us '^ •^ 

«D M ^ OO 



CO g 



I 



o 
.S 

8 



■ : &< 
& ! "S 



a 



i 

a 






-a S I I 



8 

a 
o 

i 

8 

s 

+ 5 

•o 

*S " 



o 



8 

o 
o 
B 

5 



g 

8 

a 

o. 



ii 

a a 



> > 



ill 



? 



•m •»« Pu 






o .0 



il 



o o 






^ •s ^ ja^ 

•— ^ ""^ •"• w 

© eg © « > 









s 






a 

8 

t 



< 



io8 Estimation of Inorganic Phosphoric Acid 



TABLE 4. 

Results on standard potassium iodide obtained with Schirmer's urea-nitrite 

method. 



WITBOUT PICRIC ACID 



WITH PICRIC ACID 



KI solution used, cc \ 25.00 25.00 25.00 25.00| 25.001 25.00 

Thiosulphate required, cc 8.2C 8.24 8.20 8.24', 8.30i 8.21 



Average titration. . . 
Calculated titration. 



8.24 
8.22 



8.25 
8.22 



iodine determination, and two portions of 50 cc. for the phosphoric acid 
determination by the combined magnesia mixture and Lorenz methods. 
The results are given in Table 5. 

It remained finally to check the accuracy of Modification C 
in the presence of proteid, for which purpose the following experi- 
ment was carried out. 

From a thoroughly groimd sample of beef round, six portions of about 
20 gm. each were .accurately weighed into large porcelain mortars con- 
taining about 20 gm. of acid-washed sand. These samples were then each 
ground, with the addition of a little water, to a smooth paste, transferred 
quantitatively, with the aid of about 200 cc. of water, to a glass-stoppered 
bottle of about 500 cc. capacity, and treated successively with 10 cc. of 
2.5 normal hydrochloric acid, 5 gm. of picric acid, and 25 cc. of half normal 

TABLE 6. 
Application of the whole method (Modification C) to standard P%Oi solution. 



Thiosulphate required for whole amount of 
KI used 

Thiosulphate required for 25 cc. aliquot 

Average 

Per cent of whole solution represented by 25 cc. 

Per cent of whole solution represented by 50 cc. 

Weight of precipitate from 50 cc 

Weight of P2OJ recovered from 50 cc. (average). 

Weight of P2OJ recovered from whole solution 
/o.ooa l_«3^ 
^ .~2 1 9 6 / 

Weight of PjO* added 

Per cent of added PaO* recovered 



96.50 cc. 
9.75 cc. 



96.50 cc. 

9.74 cc. 

9.745 cc. 
10.098 
20.196 



0.0962 gm. 0:0978 gm. 
0.003162 gm. 



0.01566 gm. 
0.01564 gm. 
100.11 



R. M. Chapin and W. C. Powick 



109 



potassium iodide solution. To three of the samples, 25 cc. portions of a 
solution containing 0.001564 gm. of PsOt per cc. were added, and the inor- 
ganic phosphoric acid in each sample was then determined by Modification 
C of the method that we have already described. The results of this experi- 
ment are shown in Table 6. 

TABLE 6. 
Application of Modification C to recovery of phosphoric acid added to beef. 



WITHOUT PsOi ADDITION 

2 



WITH PlOi ADDITION 



1 



KI aolution used, oc 

Thiosnlphate equivalent of above 

Volume ot extract used for iodine 
d^ermination 

Thioeulphate required for 50 cc. . . 

Fart of whole represented by 
50 cc 

Fart of whole solution repre- 
sented by 100 cc 



Charse 

Gm. of meat reprsMOited by 100 
cc. aliquot , 

Weight of phoephomolybdate pre- 
cipitate , 

PsO» factor 



Weight of FsOt in 100 oc. (total) . . 
Fer cent of "natural" PsOi in 
meat 



25.00 
110.80 

50.00 
19.06 

0.18014 



I 



25.00 
110.80 

50.00 
20.08 



25.00 
110.80 

50.00 
20.35 



25.00 
110.80 

50.00 
19.70 



0.181231 0.183661 0.1778 



0.3602& 0.362461 0.36732 0.3556 

h 



20.1618 

7.2639 

0.6253 
0.03301 



25.00 
110.80 

50.00 
20.21 

0.1824 

0.3648 



25.00 
110.80 

50.00 
22.96 

20722 

0.41444 



20.0316 I 20.0445 



7.2607 I 7.3627 

0.6210 0.6421 
0.03301 O.O33O1I 



19.9606 

7.0980 

1.0270 
0.03304 



0.02064 



0.02050 0.02119! 0.03394 



0.2841 0.2823 



Average of first set 

Weight of "natural" PsOi in 100 cc. 



0.2878 



0.2847 



I 



0.02021 



20.0095 20.3118 

7.2995 8.4180 

1.0654 1.2097 

0.03304 0.03307 



0.03521 0.04000 



0.02078 0.02396 



Weight ot added PtO» determined in 100 cc 0.01373, 0.01442 0.01604 



PsOt originally added • 0.039i0j 0.03910 03910 

Added FsOs recovered (calculated for whole) I 0.03861! 0.03954 0.03870 



Per cent of added PfOs recovered 98. 73 



101.12 



98.97 



Average percentage of recovery ' 09 . 61 



While the averages of the individual experiments described 
above, both in the control and in the experiment on meat, show 
the practically quantitative recovery of the entire amount of 
phosphoric acid added, yet the recovery was not so nearly quan- 
titative in the individual determinations, which show errors 
amounting to over 1 per cent of the total phosphoric acid deter- 
mined. When it is considered, however, that the actual experi- 



no Estimation of Inorganic Phosphoric Acid 

mental error arising from a combination of four separate analytical 
operations is naturally considerable, and that in these experiments 
such an error is multiplied by three, a final error of ± 1 per cent 
does not seem inordinately large. 

It may therefore be said that, in the absence of interfering 
substances, the above described modification leads to rapid and 
sufficiently accurate results, and for the most part seems to be a 
very desirable variation. Yet it is felt that the modification 
has not yet received a sufficiently broad test, and that it leaves 
something to be desired on account of the possibly limited range pf 
its utiUty. Thus, it was found that it could not be used upon 
cold water extracts of meat which had been preserved with thy- 
mol, on account of the interfering action of that substance, unless 
the thymol had been previously expelled. 

Comparison of the improved method with other well known methods. 

Although the data already obtained had convinced us of the 
accuracy of our method, it was nevertheless decided to compare 
it with a few other well known methods, even though in the event 
of their yielding discrepant results, no conclusion could be formed 
as to which method might be at fault. The methods selected for 
this purpose were the Emmett-Grindley, the Siegfried-Singe- 
wald, and the Forbes methods. These methods as outlined by 
Grindley and Ross" were followed in all essential respects, except 
that after the initial phosphomolybdate precipitates had been 
formed according to directions, the subsequent procedure in all 
the methods was identical and consisted of a magnesia mixture 
precipitation followed by a final estimation of the phosphoric acid 
by the Lorenz method. 

The material selected as sample was a cold water extract of 
flesh, prepared with thjrmol as a preservative; and for the pur- 
pose of comparing the methods, the two following experiments 
were made. 

Experiment A. Two liters of a cold water extract were prepared from 
practically fresh, finely ground beef round and water saturated with thy- 
mol. Triplicate determinations were made by each of the three methods 



" Grindlev and Ross: loc, cii. 



R. M. Chapin and W. C. Powick 1 1 1 

mentioned, and by the picric acid method, 200 cc. of the extract being used 
for the picric acid treatment, and 100 cc. for each of the other methods. The 
total volume of the solution in the picric acid method was determined, after 
expelling the thymol, by Modification C. The following results were 
obtained: 



MXTHOD 



PER CENT OF SOLUBLE INOROANIC PsOl IN 
ORIGINAL MEAT 



! Results of separate detenninations Average 



Forbes magnesia mixture | 0.2856 , 0.2848 

Emmett-Grindley i 0.2792 0.3224* 

Siegfried-Singewald i 0.2812 ' 0.2800 

Picric acid 0.2750 i 0.2746 



0.2838 ' 0.2848 

0.2794 ' 0.2793 

0.2806 I 0.2806 

0.2740 0.2745 



"Rejected. 



Experiment B, The previous experiment was repeated, except that the 
meat used for the preparation of the extract was somewhat older, and that 
Modification B instead of Modification C was used for the picric acid 
method. In using Modification B the picric acid was added after the 
solution had been brought to definite volume, and no correction for volume 
was then found to be necessary. The results obtained in this experiment 
are given below: 



I PER CENT or SOLUBLE INORQANIC PtOl IN 

■ J.. , ORIGINAL MEAT 

I Results of separate determiDations I Average 



Forbes magnesia mixture 0.3519 I 0.3533 



Emmett-Grindley 0.3531 

Siegfried-Singewald. | 0.3506 



0.3536 
0.3522 



0.3523 
0.3546 
0.3487 



Picricacid ! 0.3463! 0.3459 l 0.3471 



0.3525 
0.3538 
0.3505 
0.3464 



In each of the above experiments lower results were obtained 
by the picric acid method than by any of the other methods used. 
It is believed, however, that the results of the determinations 
by our method are quantitative, since the method is theoretically 
correct and has been indicated to be accurate by the experiments 
already recorded and by numerous others which we have not 
brought forward. The larger results obtained by the other 
methods are to be explained, we believe, by the precipitation of 
some of the organically combined phosphorus, or by the Uberation 
of inorganic phosphorus from organic phosphorus compounds, or 
both. 



112 Estimation of Inorganic Phosphoric Acid 

The distribution of phosphorus in eggs, 

A few hitherto impubUshed results obtained by the use of the 
picric acid method on eggs are here appended. It was our object 
in these experiments to determine the ratio between the inorganic 
and the total phosphorus in different grades of eggs, in order to 
see if there might exist any relation between this ratio and the 
degree of deterioration as judged by candUng. The experiments 
showed a progressive increase in this ratio as deterioration ad- 
vanced. The method used for the determination of the inorganic 
phosphoric acid was the above proposed picric acid method, 
Modification A being employed. The total phosphoric acid was 
determined by igniting the dried egg substance with soda and 
saltpeter, and, after dissolving the residue and filtering, by pre- 
cipitating by the Lorenz method. The results obtained are 
shown in Table 7. 

CONCLUSION. 

L An improved method for the determination of inorganic 
phosphoric acid in tissues has been devised, which, on a priori 
grounds, should yield correct results, and which has the following 
advantages over the methods customarily used: 

a. It does not involve the action of heat or of strong reagents 
upon the organic matter present in the tissue. 

b. Bacterial decomposition of the organic matter is prevented 
by the presence of picric acid, a powerful antiseptic. 

c. The precipitation is effected after the removal of the proteid 
matter, and therefore is not influenced by the protective action of 
colloids; while contamination with organic compounds of phos- 
phorus is reduced to a minimum. 

d. Tedious filtrations are avoided by using an aliquot portion 
of the filtrate from the picric acid coagulum for the determination 
of inorganic phosphoric acid. 

e. A rapid and accurate chemical method has been employed 
for determining the proportional part of the whole solution repre- 
sented by the aliquot part used. 

2. The method described, and its several modifications, have 
been experimentally tested and found to be rapid and accurate. 



R. M, Chapin and W. C. Powick 



113 



TABLE 7. 



Distribution of phosphorus in 


different grades of eggs. 




OR 

3s 


a 

1 
s 


XMORGANIC 
PlOt Ui 


TOTAL 
PlO» IN 


» « 


uEscKiniON or bamfle 


Fresh 
egg 


Dry 

sub- 
stance 

0.0683 
0.0610 
0.0612 


Fresh 
egg 


Dry 

sub- 

Btaoce 


Hit 


Fresh cfss, 24 hfB. old 


49 




0.0165 
0.0159 
0.0159 










1 
73.98, 0.0161 


0.0618 








Fresh cfcs, 24 hn. old 


50 




0.0179 
0.0179 
0.0183 
0.0182 
0.0183 


0.0689 
0.0686 
0.0704 
0.0703 
0.0704 


0.494 


1.900 




. 


73.98 


0.0181 0.0697 


0.494 

0.493 
0.496 

0.495 


1.900 


3.670 


Fresh eggv. 24 hrs. old 


55 


74.37 
74.32 


0.0179 

0.0180 



0.0180 


0.0700 
0.0703 


1.922 
1.935 




^ 


74.34 


0.0701 


1.929 


3.555 


*'FrMh" efcs, aTJIudfed by candling test 


52 


73.71 
73.67 


0.0156 


0.0592 

0.0592 

0.0830 
0.0838 


0.510 
0.507 


1.939 
1.928 

1.933 






73.69 


0.0156 

0.0227 
0.0229 

0.0228 


0.509 

0.543 
0.546 

0.545 

0.490 
0.492 

0.491 

0.627 
0.619 


3.064 


Second grmde eggs, k«pt froaen for 185 dys. 


57 


72.72 
72.67 


1.988 
2.002 






72.70 


0.0834 


1.995 


4.181 


Poorer grmde of No. 2 egss kept frosen for 
187 dys. 


58 


69.67 
69.38 


0.0207 
0.0209 


0.0687 
0.0679 


1.607 
1.615 

1.6II 

2.392 
2.360 






69.52 


0.0208 

0.0279 
0.0279 

0.0279 


0.0683 

0.1062 
0.1063 

0.1063 


4.239 


"Borderline," almoet inedible eggs, kept 
frosen for 187 dys. 


59 


74.02 
73.52 






73.77 


0.622 


2.376 

1.879 
1.886 

1.882 


4.437 


Shell eggs, containing dead, 2 dy. old embryo; 
in incubator two wks. 


54 


70.67 
70.72 

70.70 


0.0408 
0.0418 




0.550 
0.553 






0.0413 


0.1400 


0.552 

0.536 
0.546 


7.486 


**Rots,** eggs in a well advanced stage of de- 
eompoaition 


23 


68.93 


0.1294 
0.1280 


0.4163 
0.4119 

0.4141 


1.725 
1.758 






68.93 


0.1287 


0.541' 1.741 

1 


23.78 



TBB JOUBMAL OF BIOLOGICAL CHXiaaTRT, VOL. XX, NO. 2 



I 



114 Estimation of Inorganic Phosphoric Acid 

3. Experiments for the comparison of the picric acid method 
with the Forbes, the Emmett-Grindley, and the Siegfried-Singe- 
wald methods have shown that approximately the same results 
are obtained by each, when employed for the determination of 
inorganic phosphorus in the cold water extract of flesh. The 
slightly lower results obtained by the picric acid method are 
attributed to the sharper separation of the inorganic from the 
organic phosphorus compounds that can be obtained by its use. 

4. By means of this method, a progressive increase in the ratio 
of the inorganic to the total phosphorus in eggs has been found, 
which increase corresponds to the increased deterioration of the 
eggs as judged by physical means. 



ON THE SIZE AND COMPOSITION OF THE 

THYMUS GLAND. 

By FREDERIC FENCER. 

(From the Research Laboratory in Organotherapeuiics of Armour and 

Company f Chicago.) 

(Received for publication, December 17, 1914.) 

There is abundant evidence to the effect that the thymus is most 
active during the growth period before puberty, and it is generally 
assumed that the gland becomes inactive and gradually atrophies 
after puberty. It would, therefore, be logical to use the glands 
from healthy, young, growing animals for medicinal purposes, and 
in the Armour Laboratory these glands are obtained exclusively 
from suckling calves. 

For the purpose of obtaining, if possible, some definite and con- 
clusive information regarding the relative size, physiological 
acti\4ty, and chemical composition of the thymus during intra- 
uterine life as well as before and after puberty, this investigation 
was carried out on glands from beef fetuses six to nine months 
old, from calves two to four months old, and from full grown 
cattle. Glands from adult hogs were also collected for com- 
parative purposes. Sheep glands were obtained indiscriminately, 
from young and from full grown animals, and the entire lot repre- 
sents, therefore, all ages of this species with the exception of new- 
born lambs and sucklings. It was impossible to secure glands 
from hog and sheep fetuses and from suckling pigs and lambs 
in sufficient quantity to be of much, if any, value. 

The glands were collected every Tuesday over a period of ten 
weeks, during June, July, and August, 1914. This time was 
selected because it is at the height of the outdoor season, and the 
animals are allowed, at least to a large extent, to choose the qual- 
ity and quantity of their individual food. In the winter more or 
less artificial feeding conditions prevail. 

The glands were removed from the various animals unmediately 
after slaughter, while still retaining the animal heat, carefully 

"5 



\ 



1 16 Size and Composition of the Thymus Gland 

trimmed free from comiective and other adherent tissues, weighed, 
and stored at freezing temperature until the entire lot was col- 
lected. They were then finely minced, well mixed, and average 
portions of about 2000 grams dried on enamelled trays to con- 
stant weight at a temperature not exceeding 50''C. The dried 
material was coarsely ground and extracted with petroleum ether 
in Soxhlet extractors; and the desiccated fat-free material was 
powdered in a steel tube mill to pass a 60 mesh sieve. 

In the tabulation will be found the total number of glands 
employed, the maximum, minimum, and average weights of the 
fresh glands together with the moisture, petroleum ether-«oluble 
Eubstanc ■, and yield of desiccated fat-free material. 

The averf^ weights of the live animals are also given. The 
dry, fat-free gland was taken as a basis for obtaining the exact 
proportion of active thymus tissue to the live weight of the ani- 
mals. The calculated figures in the tabulation express the milli- 
grams of dried fat-free gland tissue per kilo of body-weight, and 
demonstrate clearly that the proportion of thymus tissue is far 
larger in fetuses and in young growii^ animals than in fully mature 
animals. 

The diet apparently has some influence on the size of the thymus 
gland, as it is considerably larger in the two species of herbivorous 
animals than in the omnivorous hog. 

It was noticed that the fetal thymus glands were exceptionally 
rich in blood. 



' A ■ 






111 n\-A K I iiiistdi^i 



;jw iprr ijwr IpfT I 
.0,T9.ll 1. Ilia. 716. fll^iT. 

.ON.e{u.zis.sis.g{ «7. 

-'■■tiU.lllB.SlS.i 12. 

« 4 iwrnf, I a 



3.tl< 8. Si 11 .» in 



Frederic Fenger 117 

"1 he tabulated figures show very little difference in the moisture 
content between the fetal and calf thymus, and both lots are 
practically free from fat. The glands from full grown cattle and 
hogs are somewhat lower in moisture content, but show quite a 
high percentage of fat. The formation of adipose tissue in the 
interlobular or vascular connective tissue, which replaces the ade- 
noid tissue containing the physiologically active constituents of 
the gland, is evidently very gradual and largely dependent on the 
age of the animal. This is borne out clearly in the case of sheep 
glands. Here we have all ages represented, and it is interesting 
to note that the moisture content is considerably higher than in 
adult cattle and hogs. The fat content is much lower than in the 
full grown animal, but is somewhat higher than in the fetal 
glands, and glands from young animals. 

That the adipose tissue simply has replaced the adenoid tissue 
is shown by the almost identical yield of desiccated fat-free ma- 
terial obtained in all three cases where the full grown animals 
predominate. 

On the desiccated, fat-free material the following determina- 
tions were made: moisture, ash, phosphoric acid, and total nitro- 
gen. The moisture was determined on 5 gram samples of the 
powdered material in flat porcelain dishes by drying at 100°C. 
to constant weight. This required five hours. The ash was de- 
termined on 1 gram samples. The powder was weighed into 
tared porcelain crucibles of 15 cc. capacity and heated at a dull 
red heat in a muffle furnace to constant weight. This required 
approximately three hours. 

The determinations of phosphorus were made according to the 
Neumann "wet" combustion method followed by titration of the 
yellow ammonium molybdate precipitate according to the details 
given in the volumetric method of the Association of Official 
Agricultural Chemists. 

The total nitrogen was determined according to the official 
Kjeldahl-Gunning-Arnold method described in Bulletin 107, 
Bureau of Chemistry. 

From the tabulated results it will be seen that the ash content 
of the thymus gland is high and consists principally of phosphates. 
The ash as well as the phosphorus contents are highest in glands 
from fetuses and very young animals, which, of course, strengthens 



ii8 Size and Composition of the Thymus Gland 

the conclusion that the gland is most active and contains the 
highest percentage of active principle before puberty. There is 
but sUght difference between the ash and phosphorus content of 
the fetal glands and glands from young calves. This indicates 
that the gland contains active principles and is therapeutically 
active during intra-uterine life. 

If the figures found for total nitrogen are multiplied by the 
protein factor 6.25, it will be seen that the sum of the ash and pro- 
tein is considerably above 100 per cent in all instances. This, 
of course, depends upon the fact that the thymus is very rich in 
nuclein- bodies and relatively poor in ordinary proteids. It is 
interesting to note that the fetal glands also contain these nuclein 
bodies in quantities equkl to those present in the glands from young 
animals. It appears, therefore, that the fetal thymus is in full 
activity at least three months before maturity. 

Since the nuclein bodies are present in glands from full grown 
animals also, it is reasonable to assume that this gland does not 
cease its systemic activity throughout the growing and repro- 
ductive periods of the life of these animals. 

SUBIMARY. 

The fetuses and young growing animals contain very much more 
thymus tissue per unit of body-weight than fully mature animals. 

The fetal thymus is exceptionally rich in blood and contains 
nuclein bodies and phosphates in amounts equal to those found in 
the glands from young growing animals, indicating that the gland 
is active therapeutically at least three months before maturity 
of the fetus. 

Thymus glands from full grown animals also contain nuclein 
)x)dies and phosphates, indicating that the gland does not com- 
pletely cease its systemic activity during the reproductive period 
of these animals. 

Adult cattle and sheep (herbivora) contain more thymus tissue 
per unit of body-weight than adult hogs (omnivora). 



THE INFLUENCE OF A DIET OF MARINE ALGAE UPON 
THE IODINE CONTENT OF SHEEP'S THYROID. 

By ANDREW HUNTER and SUTHERLAND SIMPSON. 

(From the Department of Physiology and Biochemistry^ Cornell University, 

Ithaca.) 

(Received for publication, December 21, 1914.) 

It is a well known fact, emphasized afresh by each new series 
of analyses, that the thyroid gland, in respect to its iodine con- 
centration, is subject to notable variations. This is true not 
•merely for different species, but also for individuals within the 
species, and even for the same individual at different times. 
Among the factors responsible for these variations it is generally 
believed that the iodine content of the food plays the most im- 
portant part. If this opinion be well founded, it would be ex- 
pected that any group of animals subsisting upon a diet unusually 
rich in iodine should exhibit an unusual concentration of that 
element within its thyroids. We have recently, in the case of a 
group of sheep living under rather peculiar conditions, had an 
excellent opportunity of putting this expectation to the test; 
and we have found it, in that instance at least, to be fully re- 
alized. The observation seems to us of sufficient interest to te 
placed on record. 

In many of the islands of the Orkney group, which lies to the 
north of Scotland, the native sheep run wild. They are of a 
small and hardy breed and during the winter months subsist, to a 
very large extent, on seaweed. Our material was obtained from 
one of these islands in the month of December. When the sheep 
were killed the thyroids were removed by a farmer, under the 
direction of the resident medical officer, and inmiediately trans- 
ferred to 95 per cent alcohol, the gland of each animal (both lobes) 
being placed in a separate bottle. No record was kept of the body- 
woight, age, or sex of the animals, but all were adults. The bot- 

119 



120 Influence of Marine Algae on Sheep's Thyroid 

ties were sealed with paraffin to prevent evaporation, and the 
material was received by us, in good condition, about a month after 
it had been collected. 

Each specimen received was removed from the alcohol in which 
it was preserved, thoroughly dried, and reduced to powder. Its 
iodine content was determined by the method described by one of 
us four years ago.^ Whenever the amount of material permitted, 
the analysis was made in duplicate. The results are shown in 
Table I, where the glands are arranged in the order of their per- 
centage iodine content. 

TABLE I. 



NO. 
1 

2 
3 

4 

« 

7 
S 

9 
10 



DRY WEIGHT 
or GlJiSV 



AMOUNT TAKEN 
FOB AKALTBIR 



IODINE POUND I IODINE FOUND 



gm. 

0.45* 
0.54 

1.25 
0.81 
1.15 

1.15 

0.6C 
0.97 

1.37 
0.94 



gm. 

0.428 
0.527 
0.408 
0.768 
0.387 
0.369 
0.613 
0.485 
0.451 
0.636 
0.654 
0.546 
0.646 
0.681 
0.420 
0.442 



mgm. 

1.79 
2.63 
2.23 
4.31 
2.37 
2.19 
4.16 
3.24 
3.19 
4.52 
5.15 
4.72 
5.96 
6.40 
4.35 
4.70 



per cent 



0.547 
0.561 
0.612 
0.594 
0.679 
0.670 
0.707 
0.711 



0.923 
0.940 
1.037 
1.064 



} 
} 
} 
} 

} 
} 



0.418 
0.499 

0.554 
0.603 
0.674 

0.709 

0.788 
0.864 

0.931 
1.050 



TOTAL IODINE 
or OI.AND 

mgm. 



2.70 
6.94 

4.89 

7.74 

8.15 

5.20 
8.37 

12.77 
9.87 



* One lobe only. 



It appears from this table that in the glands under consid- 
eration the concentration of iodine ranges from 0.42 to 1.05 per 
cent of the dry substance, the average for the whole group of ten 
being as high as 0.709 per cent. The remarkable character of 
these figures is best revealed by comparison with the recent 
analyses of sheep's thyroid, which we have brought together in 
Table II. 



* A. Hunter: this Journal, vii, p. 321, 1909-10. 



Andrew Hunter and Sutherland Simpson 121 



TABLE II. 



LOC ALITY 



MO. or 

QLANDS 



Ohio 

Western New 

York 

Western New 

York 

Western New 
York 

Xewcastle-on- 
Tvne 

Edinburgh 



19 

10 

12 
Very large 

Very large 

Very large 
Very large 



Range or 

IODINE 
CONTENT 



percent 



AVERAOE 

IODINE 

CONTENT 



percent 



0.125-0.461 0.247 



0.048-0.383 

0.249-0.580 
0.049-0.260 

0.048-0.335 

0.280-0.510 
0.222-0.279 



0.168 

0.407 
0.158 

0.167 

0.375 
0.251 



OBSERVER 



Marine and Lenhart.* 

Simpson and Hunter.' 

Simpson and Hunter.* 
Seidell and Fenger.* 

Seidell and Fenger.* 

Martin.^ 
Guyer. * 



The last four sets of figures in Table II are derived from 
analyses, not of individual glands, but of composite powders 
obtained from groups of many hundreds. When, therefore, 
Martin reports an iodine content of 0.51 per cent, it is clear that 
his material must have included some glands that had an even 
higher percentage of the element. The figure 0.58 per cent ob- 
served by us in a sheep from Western New York could probably, 
therefore, be duplicated, if not surpassed, among sheep from 
Northern England. As it stands, however, it is the maximum 
value for sheep's thyroid hitherto recorded. It approaches 
closely the highest concentration of iodine thus far observed in 
any mammalian thyroid whatsoever; namely, 0.629 per cent for 
a canine thyroid analyzed by Marine and Lenhart.® Yet of the 
Orkney thyroids reported in this paper, no fewer than seven 

» D. Marine and C. H. Lenhart: Arch, Inl, Med,^ iv, p. 440, 1909. 
'S. Simpson and A. Hunter: Quart. Jour. Exper. Physiol., Hi, p. 121, 
1910. 

* Simpson and Hunter: ibid., iv, p. 257, 1911. 

» A. Seidell and F. Fenger: Bull. Hyg. Lab., U. S. P. H. and M.-H. S., 
Xo. 96, p. 67, 1914. 

• Seidell and Fenger: this Journal, xiii, p. 517, 1912-13. 

^ Martin: Pkarm. Jour., Ixxxix, p. 144, 1912, and xci, p. 126, 1913. Quoted 
from Seidell and Fenger: loc. cit., 1914. 

• Guyer: Pkarm. Jour., xci, p. 123, 1913. Quoted from Seidell and Fen- 
ger: loc. cit. 

* Marine and Lenhart^ loc. cit. 



122 Influence of Marine Algae on Sheep's Thyroid 

surpass the sheep, and six the mammalian record. Even the 
average for this group exceeds quite notably the previous maxhna ; 
while the series is closed by a gland containing the really aston- 
ishing quantity of over 1 per cent of iodine. 

There is in the literature but one instance in which this last 
particular gland has been surpassed. It is supplied by Cameron^° 
who found 1.16 per cent of iodine in a composite sample of 
thyroid powder from the elasmobranch fish, ScyUium canicula. 
That fish thyroids should frequently contain unusually large 
amounts of iodine is of particular interest in the present connec- 
tion. Cameron connects his observation quite naturally with the 
constant presence of iodine in sea water; and in an extensive re- 
view of the distribution of iodine in plant and animal tissues," 
he concludes that the variations in the iodine content of thyroid 
tissue are all referable to differences in diet. Without denying 
that other factors occasionally play a part," we take our present 
results to be confirmatory of the general correctness of this view. 
While we have not identified nor analyzed the seaweeds consumed 
by our group of animals, there can be no doubt that they con- 
tained iodine.*' That the sheep, like other animals, can store 
in its thyroid the iodine of its food is likewise beyond question; 
we have ourselves observed,** among other instances, one where 
the continued daily administration for six or seven weeks of small 
quantities of sodium iodide raised the iodine content of a sheep's 
thyroid from 0.53 to 1.15 per cent. There seems, therefore, to 
be no occasion to look beyond the food supply for an explanation 
of the analytical data we have presented. 

*® A. T. Cameron: Biochem. Jour.f vii, p. 466, 1913. 

" Cameron: this Journal y xviii, p. 335, 1914. 

^2 A discussion of factors accounting for seasonal variations will be 
found in Seidell and Fenger: loc. cit., 1914. 

" Our existing information upon the iodine content of different seaweeds 
is detailed by Cameron: this Journal^ loc. cit. 

^* Simpson and Hunter: Quart. Jour. Exper. Physiol. , loc. cit. 



A NOTE ON THE DISTRIBUTION OF MERCURY IN THE 

BODY IN A CASE OF ACUTE BICHLORIDE 

OF MERCURY POISONING. 

By JACOB ROSENBLOOM. 

{From the Biochemical Laboratory of the Western Pennsylvania Hospital, 

Pittsburgh.) 

(Received for publication, December 21, 1914.) 

The writer recently assisted in an autopsy on the body of a young 
girl who had, ten days previously, taken about 15 grams of bi- 
chloride of mercury for suicidal purposes. The autopsy was 
held about five hours after death, apd it was thought that it 
would be of value to estimate the amount of mercury contained 



ORGAN 


MERCURY IN* MOM. 

PER 100 OM. OP 

ORGAN 

2.80 

1.20 

3.62 

0.14 

0.50 

1.60 

1.82 

1.92 
Trace 

1.20 

0.06 
Trace 


MERCURY IN MOM. 

IN WHOLE 

ORGAN 

1 


Kidney 


5.80 


Spleen 


1.70 


Liver 


39.82 


Brain 

Stomach 


1.68 
1.50 


Small intestine 


' 4.80 


Larse intestine 


7.28 


Heart 


5.81 


Lunes 


Trace 


s^ 

Blood 


52.0* 


Muscle 


25.8+ 


Bile 


Trace 


Stomach -contents 


12.20 


Intestinal contents 


S.OO 


Rectal contents 


S.40 






Total 


175.39 







* Calculated od the basis of the blood constituting ,^ of the I >oiiy- weight, 
t Calculated on the basis of the muscle constituting | of the body-w eight. 



123 



1 24 Distribution of Mercury in the Body 

in various organs of the body.^ The mercury content of the 
organs was estimated by Ludwig's method.^ 

The above tabulated data give the amounts of mercury found 
in the various organs and fluids of the body. They show that 
the liver contains the largest amount of mercury per 100 grams 
of tissue, but that the blood contains the largest quantity based 
on the whole amount present in the body. 

The mercury found in the contents of the stomach, intestines, 
and rectum may, of course, represent mercury absorbed and 
regxcreted into those channels. 

* A thorough review of the literature on the distribution of mercury in 
the body may be found in Witthaus and Becker: Medical Jurisprudejice, 
Forensic Medicine and Toxicology , iv, p. 721, 1911. 

* E. Ludwig: Ztschr. ]. anal, Chem,^ xvii, p. 395, 1878 (cited by C. Neu- 
bauer); xx, p. 475, 1881 (cited by F. Hofmeist^r); Wien, med, Presse, 
xxxiii, p. 1891, 1892; £. Ludwig and E. Zillner: Wien. klin. Wchnschr., ii, 
p. 858, 1889; iii, p. 534, 1890. 



ON THE ESTIMATION OF BENZOIC ACID IN URINE. 

By G. W. RAIZISS and H. DUBIN. 

(From the Department of Dermatological Research^ Philadelphia Polyclinic 

and College for Graduates in Medicine y Philadelphia^ Pa., 

Dr. Jay F, Schambergj Director.) 

(Received for publication, December 28, 1914.) 

For the quantitative estimation of free benzoic acid in the urine 
of rabbits we have found that none of the methods described in 
the literature are entirely satisfactory in the points of convenience 
and rapidity. Recently, Steenbock^ has reviewed the methods 
used for the determination of benzoic and hippuric acids, recom- 
mending Dakin's method for isolating benzoic acid, with further 
purification by sublimmg according to his own modification. Soon 
afterwards, Folin and Flanders^ published their method for the 
determination of hippuric acid, which is, because of its conven- 
ience, superior to any previously described. According to this 
method, the hippuric acid is hydrolyzed into benzoic acid and 
glycocoll by first evaporating the urine with alkali on the water 
bath, and then boiling for several hours with nitric acid and 
copper nitrate. The resulting benzoic acid is extracted with 
chloroform and titrated with -^ sodium alcoholate. 

The extraction of free benzoic acid from fresh urine with chloro- 
form would be feasible, if it were not for the troublesome emul- 
sions formed. Working with the urine of rabbits which had 
received large amounts of benzoic acid, we could get no satis- 
factory results, chiefly because of the great extent to which the 
chloroform emulsified. Heating or any other treatment of the 
urine with the purpose of removing the emulsifying substances, 
resulted in splitting of hippuric acid or the loss of benzoic acid. 

Consequently we have sought a solvent of benzoic acid which 
would form the least emulsion when shaken with fresh urine. 

* H. Steenbock: this Journal^ xi, p. 201, 1912. 

* O. Folin and F. F. Flanders: this Journal, xi, p. 257, 1912. 

125 



126 Benzoic Acid in Urine 

Out of many solvents tried, toluene appeared to be the best. 
While it is an excellent solvent for benzoic acid, it practically 
does not dissolve hippuric acid, and is superior to chloroform 
in many ways.' As a rule, toluene does not produce emulsions 
when shaken with fresh urines, but small emulsions may form 
when working with abnormal urines; for example, those contami- 
nated with feces. In such cases it was found that the emulsion 
could be removed by shaking with a small amount of absolute 
alcohol. 

Our method is a modification of that proposed by Folin and 
Flanders* for ketchup. The procedure is as follows: 100 cc. of 
fresh urine are pipetted into a 500 cc. short stemmed* separatory 
funnel, and acidified with 1 cc. of concentrated nitric acid. Enough 
ammonium sulphate (50 to 60 gm.) to saturate the urine is added, 
and the benzoic acid is extracted with four portions of pure 
toluene* of 50, 40, 30, and 30 cc. each, respectively. The com- 
bined toluene extracts are then washed twice, using 100 cc. each 
time, with saturated sodium chloride solution, containing in each 
liter 0.5 cc. of concentrated hydrochloric acid. The titration is 
made with ^ or ^ sodium alcoholate,^ using phenolphthalein 
as an indicator. The end-point is a definite pink, lasting two 
or three minutes. 

To secure the best results, the following points should be ob- 
served. The contents of the funnel are shaken with a rotary 
motion, from 15 to 20 times with each portion of toluene. If 
the shaking is not too vigorous, no emulsion should occur. Should 
there be a sUght emulsion, however, it is disregarded for the 
present, and obviated in a later stage of the analysis by addition 
of absolute alcohol. After the first extraction the urine is drawn 
off into a second funnel, leaving behind the toluene and any 

' One of the advantages of toluene is that it does away with the dis- 
comforts so often experienced while working with chloroform. 

* Folin and Flanders: Jour, Am. Chem, Soc.f xxxiii, p. 1622, 1911. 

* The stem of an ordinary separatory funnel is cut off about an inch 
from the stop-cock. 

* The toluene used for extractions may be distilled and washed twice 
with distilled water, when it is again ready for use. 

■ Sodium alcoholate is prepared by dissolving 2.3 gm. of clean metallic 
sodium in I or 2 liters, respectively, of absolute alcohol. This solution 
is standardized against pure benzoic acid dissolved in toluene. 



G. W, Raiziss and H. Dubin 127 

emulsion that may have formed; the urine is then shaken with 
the second portion of the toluene. 

In order to avoid the use of a third funnel at this point, the 
urine is drawn ofif into an Erlenmeyer flask, and the toluene run 
into the funnel containing the first extract. The urine is then 
transferred back again into the second funnel, washing the flask 
with the required amount of toluene for the third extraction. 
The fourth extraction is carried out in a similar manner. 

Before washing the combined extracts, contained in the first 
funnel, with sodium chloride solution, any urine that may have 
collected is drawn ofif. If there is emulsion present, 1 cc. of 
absolute ethyl alcohol is added, and the funnel, held in an up- 
right position, is rotated vigorously. Drops of urine and solid 
matter settle to the bottom and are drawn off. After washing 
twice ijvath sodium chloride solution, the funnel is again rotated 
vigorously for the reason just mentioned. The stem of the funnel 
is now washed with water, dried with a piece of filter paper, and 
the toluene drawn off into a dry flask ard titrated. 

In order to show that toluene extracts benzoic acid from watery 
solution quantitatively, 0.050 gm. and 0.100 gm. of pure benzoic 
acid were dissolved by means of sodium hydrate in 100 cc. of 
water. The titrations obtained with sodium alcoholate (1 cc. = 
0.0161 gm. of benzoic acid) were 3.00 cc. and 3.05 cc. in one 
case, and 6.05 cc. and 6.10 cc. in the other. Hence the amount 
of benzoic acid recovered was 0.049 gm. and 0.098 gm., respectively. 

Four analyses of human urine, made according to the method 
described, showed the absence of benzoic acid, thus confirming 
the findings of others, and also indicating that toluene does not 
extract any substances from the urine, except benzoic acid, which 
would neutralize sodium alcoholate. 

In the accompanying table the results of a number of experi- 
ments on rabbit urines are recorded. In Column 1 the amount 
of benzoic acid found in the urine (preformed) is given. Column 
2 records the amount of benzoic acid added to the different urines. 
Column 3 shows the amount of total free benzoic acid found, 
and Colunm 4 gives the amount of added benzoic acid which 
was recovered. In Column 5 the recovered benzoic acid is cal- 
culated on a percentage basis. From the figures we see that the 
amount of benzoic acid recovered is practically quantitative. To 



128 



Benzoic Acid in Urine 



some of these urines 0.100 gm. of hippuric acid was added, in 
addition to the benzoic acid, without affecting the result, show- 
ing that none of the hippuric acid was extracted by toluene. 



(i) 



TABIjE I. 

C2) (3) 



(4) 



(5) 



(6) 



URIKE 



A.. 

B.. 
C*. 



p a o 

Q h O 
BOA 



oS 



« 



#< V K 

H » CQ 



gm. 

0.0021 
0.0018 
0.0019 
0.0065 
0.0072 
0.0020 
0.0912 
0.0330 
0.0338 



n 



am. 



0.0423 

0.0423 

0.0423 

0.0846 

0.0846 

0.100 

0.100 

0.1046 

0.1060 



o 

D 

o 

Q 

o 



z 
OS 

* r* O 

5SS 

CQ 



gm. 

0.0443 
0.0444 
0.0449 
0.0887 
0.0894 
0.1900 
0.1885 
0.1368 
0.1377 




< 



as 

0. M 

- < 
X 



gm. 



0.100 
0.100 
0.100 
0.100 



* This rabbit bad received larse amounts of benicic acid per ot. 



STUDIES ON THE THEORY OF DIABETES. 

IV. THE PARALLELISM BETWEEN THE EFFECTS OF THE PAN- 
CREAS AND THOSE OF METALLIC HYDROXIDES ON SUGARS. 

By R. T. WOODYATT. 

{From the Otko S. A. Sprague Memorial Institute Laboratory for Clinical 

Research, Hush Medical College, Chicago.) 

(Received for publication, December 28, 1914.) 

When in any case of pancreas or phlorhizin diabetes any sugar, 
such as galactose, levulose, mannose, etc., is given under condi- 
tions so selected that the sugar almost wholly escapes oxidation, 
such sugar may still be fully transformed into d-glucose. This phe- 
nomenon serves to demarcate two distinct types of sugar reaction 
in the body: the destructive and the transformative. It is only the 
former of these which disappears in diabetes. In a study of the 
fundamental chemical nature of diabetes it is important to discrim- 
inate between these and to ascertain, if possible, the chemical na- 
ture of each and the laws which govern them. Some light is 
thrown on the subject by a consideration of certain purely chemi- 
cal properties of sugars. 

Sugars in alkaline solutions in vitro manifest the same two types 
of reaction, and in this field the phenomena have been extensively 
investigated and illuminated. It has long been known that when 
a solution of any monosaccharose, such as d-glucose, is alkalinized, 
the rotatory power of the solution falls and finally disappears, the 
solution then containing a mixture of hexoses in a state of dynamic 
chemical equilibrium as shown by Lobry de Bruyn and van Eken- 
stein.^ Knowledge of the probable number and character of the 
substances participating in such an equilibrium has been extended 
by Nef.* If the conditions of the experiment are suitable (very 
dilute alkali), there need be no browning of the solution nor any 

' Lobry de Bruyn and van Ekenstein: Rec. trav. chim. de Pays Bas, xiv, 
pp. 158, 203; xv, p. 92; xvi, p. 257; xix, pp. 1, 10. 
* J. U. Nef : Ann. d. Chem., ccclvii, p. 214, 1907. 

I2Q 
«■■ JOUMKJkL or BXOLOOICAL CHBlfXaTBT, VOL. XX, NO. 2 



130 Theory of Diabetes 

detectable loss of total reducing power whatsoever. These rectp- 
rocal transformations without loss of reducing power mil be referred 
to as reactions of type I. With stronger concentrations of alkali, 
loss of reducing power occurs. In the absence of oxygen or oxidiz- 
ing agents this loss is associated with the formation of lactic acid, 
and the saccharinic acids, and browning (tars). In the presence of 
sufficient oxygen (air, H2O2, metallic oxides, etc.), oxidation prod- 
ucts occur instead of the former. These destructive reactions will 
be referred to as type IL According to Michaelis and Rona,' 
loss of reducing power was first seen in 1 to 5 per cent glucose solu- 
tions (with air excluded) when the OH ion concentration corre- 
sponded to (H • ) = 10"*^ or 10"**; whereas, in a certain instance, 
loss of rotatory power was seen with (H • ) = 7 X 10"*^ In other 
words, reactions of type I may prevail with a certain low OH ion 
concentration, those of type II appearing only with higher OH ion 
concentrations. It is of interest to inquire how a metallic hydrox- 
ide acting as a catalytic agent produces these effects. 

Sugars are weak acids. They form salts with metals and 
influence the electro conductivity of weak electrolytes as acids 
should. The dissociation constants of several sugars have been 
measured by electro chemical methods. Cohen* found for the 
dissociation constant of rf-glucoseas an acid 6 X 10~". Michaelis 
and Rona found 5 X 10-*^ In 1909 A. P. Mathews^ noted that 
the speed of oxidation of sugars in the presence of alkali was pro- 
portional up to a certain point with the concentration of the alkali, 
and proposed that metallic hydroxides acting after the manner of 
strong bases on weak acids in general formed salts with high elec- 
trolytic dissociation constants leading to an increased concentra- 
tion of sugar anions. These by virtue of a partially unbalanced 
charge are unstable and readily rearrange, oxidize, etc. Michaelis 
and Rona demonstrated that the rate of decline of rotatory power 
in glucose solutions was proportional to the OH ion concentration. 
They proposed that, in accordance with mass laws, the rising OH 
ion concentration increases the play between sugar molecules and 
ions as an incident to which new molecules appear; such, for 

* L. Michaelis and P. Uona: Biochem. Ztschr.j xlix, p. 232, 1913; xlvii, p. 
447, 1912. 

* E. Cohen: Ztschr. f. physikal. Chem., xxxvii, p. 69, 1901. 

* A. P. Mathews: this Journal, vi, p. 3, 1909. 



R. T. Woodyatt 131 

example, as enols of the type proposed by Wohl and Neuberg. 
Xef ascribes reactions of type I to dissociations occurring primarily 
at the aldehyde or ketone groups, the double bond between carbon 
and oxygen opening after the manner of simpler aldehydes and 
ketones to form hydrates with water. These hydrates acting like 
alcohols form salts which readily dissociate with the formation of 
an olefine dienol. The mechanism in the case of glyceric aldehyde 
would be as follows: 

H H 

J,OH - CHOH - CH = 0-fKOH;lCH,OH - CHOH - C - Oiq:!CH,OH COH = COH+KOH 

OH 

-glyceric aldehyde Salt of the hydrate 1-2 triose olefine dienol. 

The 1-2 triose olefine dienol is derivable from either d- or Z-glyceric- 
aldehyde or from dihydroxy acetone, and once formed from any 
of these the hydrates of all of the three trioses may be reformed 
from the enol by the simple addition of H and OH in the different 
ways permitted by the laws of space chemistry. The metallic hy- 
droxides acting as electrolytes lead to preliminary salt formation, 
the salts then dissociating like the original substances, but to a 
far greater degree. 

For present purposes it is unnecessary to enter into a detailed 
discussion of the relative advantages of these views. It may be 
said that they agree in the essential that metallic hydroxides, 
acting as electrolytes, on electrolytically dissociated sugars, initiate 
the production of an increased concentration in the solution of 
certain dissociated fragments of sugar; that these fragments are 
highly susceptible to chemical change and are in a state of dynamic 
chemical equilibrium with the undissociated sugar molecules. It is 
clear too that secondary intramolecular changes of a character not 
identifiable as electrolytic in the usual sense, follow the initial 
electrolytic reaction in any case. Nef's methylene chemistry 
begins at this point. 

The fact that a parallelism exists between the behavior of sugars 
in the body and in an alkaline solution in the test-tube does not of 
itself prove that the reactions in each case depend on a fundamen- 
tally identical mechanism. Nevertheless there is a hint given which 
it would be folly to ignore, and if for the explanation of the labora- 
tor>' sugar reactions conceptions involving dissociation and equilib- 



132 Theory of Diabetes 

rium have been found necessary, they can hardly be left out of 
consideration in the interpretation of the phenomena of carbohy- 
drate metabolism. Moreover, the application of such chemical 
conceptions at once simplifies and broadens our view of many 
metabolic processes, and we may be permitted to sketch briefly a 
possible application ta the problem of diabetes. 

In diabetes mellitus there is an excess of sugar in the blood and 
still reactions of type II fail, while those of type I persist, and by 
excessive feedings of d-glucose the conditions are relatively little 
altered, provided the case is one of genuinely total diabetes. This 
is quite parallel to what occurs in a sugar solution in which the 
alkali is sufficiently dilute. In such cases of diabetes there is a 
lack of the '* something derived from the pancreas,'' and it is a short 
step to propose that this something does for sugar what in vitro 
is accomplished by alkali sufficiently concentrated.* According to 
this interpretation we have in diabetic blood an hyperglycemia 
only of molecular or undissociated inert sugar, but an hypogly- 
cemia as regards active dissociated sugar of the kind necessary for 
oxidative or type II reactions. 

In phlorhizin diabetes there is an absolute hypoglycemia as 
regards sugar of all forms. Here also reactions of type II disappear. 
But in this case the feeding of glucose in high concentration causes 
the reappearance of oxidative reactions, so that here the equiva- 
lent of alkali derived from the pancreas cannot be regarded as 
lacking. Is it not possible that, as a part of the general hypogly- 
cemia, the concentration of dissociation residues of the kind neces- 
sary for type II reactions is also low, and that diabetes mellitus 
and phlorhizin diabetes have in common a low blood and tissue 
concentration of certain forms of dissociated sugar? This would 
account for the identity of the metabolic sequences which follow 
the development of either kind of diabetes, provided that the under- 

• So far the study of body fluids and tissues by physico-chemical meth- 
ods has failed to reveal such concentrations of OH ions as would be neces- 
sary accompaniments of reactions of either tjrpe I or II in a simple aque- 
ous solution outside the bod v. The above statement in the text is not 
intended to imply that sugar dissociation is necessarily accomplished in 
the body by alkaline hydroxides or OH ions, but rather by those sub- 
stances or that substance which under physico-chemical conditions found 
in the cell exert on sugars an effect equivalent to that of metallic hy- 
droxides in a watery solution. 



R. T. Woodyatt 133 

lying hypoglycemia of phlorhizin diabetes is not obliterated by a 
too rapid entry of sugar into the blood from without, or by a 
blocking of the excretion of sugar through the kidneys.' 

In this connection it is always borne in mind by the writer that 
the pancreas is a member of the group of so called endocrinous 
glands, and that if its internal secretion acts as a dissociating agent 
for a certain class of metabolites it would be reasonable to expect 
that some other internal secretions behave in a similar way with 
other classes of metabolites. Thus the thyroid may be concerned 
with the dissociation of amino-acids or polypeptides derived from 
the breakdown of protein. There is evidence bearing on this sub- 
ject which cannot be given here. The possibility is clearly pre- 
sented that between certain breakdown products of fats, protein, 
and carbohydrates, common dissociation residues or systems of 
residues may exist which throw these classes of substances into 
an actual state of organic chemical equilibrium with one another, 
— a conception which bears directly on the nature of the correla- 
tion of the ductless glands and in a most tangible manner, which 
we hope to detail in later communications. This idea also bears 
on the question of how carbohydrates or fats suppress protein 
catabolism, etc. 

' In using the terms "blood** and "kidneys" the writer is not uncon- 
scious of the fact that cells in general must be able to accomplish within 
themselves that which blood and kidneys do for the body as a whole, and 
that the excretion of sugar from a tissue may be blocked without injury 
to the kidneys. Also, the blood is merely one tissue. 



CONDITION OF CASEIN AND SALTS IN MILK. 

BY LUCIUS L. VAN SLYKE and ALFRED W. BOSWORTH. 

(From the Chemiccd Laboratory of the New York AgriciUtural Experiment 

Station, Geneva.) 

(Received for publication, December 28, 1914.) 

INTRODUCTION. 

The chemistry of milk has been studied by many investigators. 
Numerous facts have been accumulated relative to the amounts 
and properties of the more prominent constituents of milk, includ- 
ing various conditions affecting the composition; but much less 
attention has been given to thorough study of individual constit- 
uents, owing largely to the difficulties involved in making such 
investigations. 

From the beginning of its existence, this Station has given much 
attention to study of different phases of the composition of milk. 
In connection with the study of the relation of the constituents 
of milk to cheese-making, to fermented beverages made from milk, 
and to the uses of milk in human nutrition, numerous chemical 
questions have constantly arisen and continue to come up, to 
which satisfactory answers can not be given, owing to our lack of 
knowledge of the chemistry of some of the milk constituents. 
Until our knowledge in this field becomes more complete, we can 
not understand fully, for example, the fundamental chemical 
facts involved in the process of cheese-making and cheese-ripening, 
the chemical changes taking place in its constituents when milk 
sours or when it is made into fermented beverages, such as kumyss, 
imitation buttermilk, matzoon, zoolak, bulgarzoon, etc. 

We have in hand investigations relating to several of the funda- 
mental questions referred to. In the present Inilletin we shall 
present the results of our work bearing on the following points: 

(1) Properties and composition of milk serum or constituents in 
solution. 

135 



V 



136 Condition of Casein and Salts in Milk 

(2) Properties and composition of portion of constituents not 
in solution. 

(3) Acidity of milk and milk serum. 

(4) The salts of milk. 

METHOD OF PREPARING MILK SERUM. 

Before taking up the detailed results relative to these lines of 
investigation, we will give a description of the method used in 
preparmg milk serum from milk. 

That p)ortion of the milk consisting of water and the compounds 
in solution is known as the milk serum. In studying the indi- 
vidual constituents of milk, it is necessary to separate the serum. 
Various methods have been used to separate milk serimi from the 
other constituents of milk, but the one best adapted for investi- 
gational purposes depends upon the fact that when milk is brought 
into contact with a porous earthenware filter, the water passes 
through, carrying with it the compounds in true solution, while 
the compounds insoluble in water, or in suspension, remain on the 
surface of the filter. In one form or another, this fact has been 
utilized in studying milk by Lehman, Duclaux, Eugling, Soldner, 
and others. The form of earthenware filter used by us is much 
superior to any employed by these investigators. We have made 
use of tlie special form of apparatus designed by Briggs^ for the 
purpose of obtaining water extracts from soils. Briefly stated, 
the process consists in putting the milk to be examined into a 
tubular chamber surrounding a Pasteur-Chamberland filterhig 
tul)e; pressure, amounting to 40 to 45 pounds per square inch, is 
applied by inesCns of a pump which forces air into the chamber 
containing the milk and causes the soluble p)ortion of the milk 
to pass through the walls of the filter from the outside to the 
inside of the filter tube, from which it runs out and is caught 
in a flask standing underneath. The insoluble residue accumu- 
lates on the outside surface of the filter tube, from which it can 
easily be i*emov;et] by light scraping. 

* L. J. Brifi^gs and M. H. Lapham: U. S. Department of Agriculture^ 
Bureau of Soils, Bull. 19, p. 31, 1902; O. Schrciner and G. H. Failyer: 
ideniy Bull. SI, pp. 12-16, 1906. 



L. L. Van Slyke and A. W. Bosworth 137 

It has l>een found by Rupp^ that the filter appears to have the 
power of adsorbing some of the constituents of the serum until a 
volume of 50 to 75 cc. has passed through, after which the filtered 
serum is constant in composition. In our work, therefore, the 
first p)ortion of serum filtered is not used. 

Before being placed in the apparatus for filtration, the milk is 
treated with some antiseptic to prevent souring during the process 
of filtration. 

The comp)osition of the solid portion of milk removed by the 
filtering tube is ascertained by difference; from the figures ob- 
tained by an analysis of the original milk we subtract the results 
of analysis given by the serum. 

Properties and composition of milk serum. 

Serum prepared from fresh milk by the method descril)ed 
above has a characteristic appearance, being of a yellow color 
with a faint greenish tinge and slight opalescence. 

The serum from fresh milk gives a slight acid reaction to phenol- 
phthalein and a strongly alkaline reaction to methyl orange. We 
will later give the result of a special study made of the cause of 
aciditv in milk serum. 

In the table below we give the results of the examination of two 
samples of fresh milk, the serum of which was prepared in the 
manner already described. These samples of milk were treated 
with chloroform at the rate of 50 cc. per 1000 cc. of milk, and the 
fat was removed by means of a centrifugal machine; the removal 
of fat is necessary since it clogs the p)ores of the filter. The fat- 
free milk was then filtered through Pasteur-Chamberland filter- 
ing tubes. Analyses were made of the milk and of the serum. 
We did not determine those constituents present in milk only in 
traces, such as iron, sulphuric acid, etc. 

A study of the data contained in Table I enables us to show the 
general relation of the constituents of milk to the constituents of 
milk serum. The following form of statement furnishes a clear 
sununary of the facts: 

* P. Rupp: U. S. Department of Agriculture^ Bureau of Animal In- 
dustry, Bull. 166, p. 9, 1913. 



138 Condition of Casein and Salts in Milk 



1. • Milk constituentfl 
in true solution 
in milk serum. 



(a) Sugar 

(b) Citric acid 

(c) Potassium 

(d) Sodium 

(e) Chlorine 



2. Milk constituents 
pertly in solu- 
tion and i>artly 
in suspension or 
colloidal solu- 
tion. 

(a) Albumin 

(b) Inorganic phosphate 

(c) Calcium 

(d) Magnesium 



TABLE I. 

Constituents of milk serum. 



3. Milk constitueuts 
entirely in sus- 
pension or col- 
loidal solution. 



(a) Fat. 

(b) Casein. 



CONSTITUBNTB 



8A1CPLK 1 



SAUPLE 2 



Sugar 

Casein 

Albumin 

Nitrogen in other compounds 

Citric acid 

Phosphorus (organic and inorganic) 

Phosphorus (inorganic) 

> Calcium 

M Magnesium 

V Potassium 

I Sodium 

("hlorine 

Ash 



}' 



J4 

s 

V 

■H 

gtn. 



3.35 
0.525 



0.125 
0.096 
0.128 

;0.012 

0.354^ 
0.061 



S 

9 



"8 



•oi.S 

&^ 

5 m fl « 

!|aiS 



gtn. 

0.00 
0.369 



0.067 
0.067 
0.045 
0.009 

0.352* 

0.062 



a 



•3 • 

|8 



0.00 
70.29 



53.60 
70.00 
35.16 
75.00 

99.44 

100.00 



gm. 

5.75 

3.07 

0.506 

0.049 

0.237 

0.067 
0.144 
0.013 
0.120 
0.055 
0.076 
0.725 



a 



8 



c 



I 



If §s 



gm. 

5.75 
0.00 



100.00 
0.00 



0.188 37.15 
0.049100.00 
0.237jl00.00 

I 
0.056!' 64.40 
0.048! 33.33 
0.007! 53.85 
0.124'100.00 
0.057100.00 
0.081100.00 
0.400 55.17 



*As chlorides. 



The behavior of niilk albumin attracts special attention on ac- 
count of marked lack of regularity in the results obtained. We 
commonly think of milk albumin as readily and completely 
soluble in water, and the question is therefore raised as to why a 
considerable portion of it does not pass through the Pasteur- 
C^hamberland filter. In \'iew of all the facts available, the most 
probable explanation that hjis so far suggested itself is that in 
fresh milk a part of tlu» albumin is held by the adsorbing power 
of casein. This suggestion is supported by results obtained in the 



L. L. Van Slyke and A. W. Bosworth 139 

following experiments: Serum was prepared from chloroformed 
fresh milk treated in different ways. In Experiment 1 serum 
direct from the fresh milk was compared with serum obtained 
from whey which had been obtained from another portion of the 
same milk by treatment with rennet extract. In Experiment 2 
serum direct from fresh milk was compared with (a) serum ob- 
tained from another portion of the same milk after souring and 
(b) serum obtained from another portion of the same milk to which 
some formaldehyde solution had been added. Albumin was 
determined in each case by boiling after addition of acetic acid, 
following the details given in the provisional method of the Asso- 
ciation of Official Agricultural Chemists. The results of the ex- 
periments are given below. 

EXPERIMENT 1. 

ALBT7MIN OF 
ALBUMIN PER MILK RECOVKRED 

100 CX:. IN SERUM 

gtn. per cent 

Fresh milk .312 

Serum from fresh milk . 143 45 .83 

Serum from whey .187 59 .94 

EXPERIMENT 2. 

Fresh milk .266 

Serum from fresh milk . 148 55 .64 

Serum from sour milk .253 95 .11 

Serum from milk plus formaldehyde .245 92 .21 

In Experiment 1 it is seen that when casein is precipitated by 
rennet solution the curd (the precipitated casein or paracasein) 
carries down part of the albumin with it; the amount thus carried 
down is approximately equal in this case to that retained along 
with the casein on the external surface of the Pasteur-Chaniberland 
filtering tube, when whole milk is filtered through such a filter. 

In Experiment 2 we see that when the casein is precipitated 
with acid, as in the case of natural souring, the adsorbing action 
of the casein is practically prevented and little or no albumin is 
carried down with it. In the case of the addition of formaldehyde 
to milk, the adsorbing power of casein is greatly diminished, 
probably due to the chemical reaction between casein and 
formaldehyde. 



140 Condition of Casein and Salts in Milk 

Properties and composition of portion of milk in suspension or 

colloidal solution. 

Some of the constituents of niilk are suspended in the form of 
solid particles in such an extremely fine state of division that they 
pass through the pores of filter paper, and they do not settle as a 
sediment on standing, but remain permanently afloat, and they 
cannot be seen except by ultramicroscopic methods. When 
substances are in such a condition, they are said to form a colloidal 
solution. In passhig milk through the Pasteur-Chambcrland 
filtering tube, the constituents in suspension as solid particles, 
that is, in colloidal solution, are retained in a solid mass on the 
outside of the tube and can therefore be readily obtained for study. 

(1) Appearance, When prepared by the method of filtration 
previously descril)ed, the insoluble portion of milk collecting on 
the outside of the filtering tube is grayish to greenish white in 
color, of a glistening, slime-like appearance, and of gelatinous con- 
sistency. When dried, without purification by treatment with 
alcohol, etc., it resembles in appearance dried white of egg. 

(2) Behavior with water. The deposit of insoluble milk con- 
stituents on the outside of the filtering tube, when removed and 
shaken vigorously in a flask with distilled water, goes into sus- 
pension, and the mixture has the opaque, white appearance of the 
original milk. The deposit is, of course, more or less mixed with 
adhering soluble constituents, but can be readily purified by shak- 
ing with distilled water and filtering several times. The purified 
material goes readily into suspension on shaking with water and, 
if treated with a preservative, will remain indefinitely without 
change other than the separation of fat globules. It has been 
held by some that the citrates of milk perform the function of 
holding the insoluble phosphates in suspension, but this is not 
supported by the behavior of the insoluble portion shown in our 
experiments. 

(3) Reaction. A suspension of the insoluble constituents of 
milk, prepared in the manner descril>ed above, is neutral to 
phenolphthalcin. We purified the deposit made from 1000 cc. 
of milk, made a suspension of it in water, and, after the addition 
of 10 cc. of neutral solution of potassium oxalate, it was found to 
require only 0.5 cc. of -iq- solution of sodimn hydroxide to make it 



L. L. Van Slyke and A. W. Bosworth 141 

neutral to phenolphthalein. We interpret this to mean that there 
are no tri-basic (alkaline) phosphates in milk or in the serum; 
because the serum, since it is acid, can contain none, and the 
insoluble portion, being neutral, can therefore contain none. 

(4) Relaticm of inorganic constituents to casein in milk. Without 
going into a detailed discussion of the history of the diflferent 
views held by investigators, it* is sufficient for our purpose to 
state that three general views have been put forv/ard in regard to 
the relation of inorganic constituents to casein in milk: (1) That 
milk casein is combined with calcium (about 1.07 per cent) to form 
a salt, calcium caseinate (which is neutral to Utmus and acid to 
phenolphthalein); (2) that casein is chemically combined directly 
with calcium phosphate; (3) that casein is a double compound 
consisting .of calcium caseinate combined with calcium phosphate. 

We have attempted to learn what is the true condition of casein 
in milk in relation to inorganic constituents, whether it is in com- 
bination with calcium alone or with some other inorganic base in 
addition, and also whether milk casein is an acid salt or a neutral 
salt, and, further, -'lether ihei >»r:5oluble phosphates are in combi- 
nation with casein or not. 

In studying the problem, we will first give the results of work 
done with sixteen samples of fresh milk from thirteen individual 
cows. The different determinations made are as follows: (a) 
casein, (b) total phosphorus, (c) soluble phosphorus, (d) insoluble 
phosphorus (b minus c), (e) insoluble organic phosphorus (casein 
multiplied by 0.0071), (f) insoluble inorganic phosphorus (d minus e), 
(g) total calcium, (h) soluble calcium, (i) insoluble calcium (g 
minus h), (j) total magnesium, (k) soluble magnesium, (1) insoluble 
magnesium (j minus k). The determinations of casein, total 
phosphorus, total calcium, and total magnesium were made with 
the normal or whole milk, while those of soluble phosphorus, 
soluble calcium, and soluble magnesium were made with the serum 
obtained by filtering through Pasteur-Chamberland filtering tubes 
in the manner ahready described. The amount of organic phos- 
phorus was foimd by multiplying the percentage of casehi by 0.0071.' 
For convenience of reference, the analytical data are arranged in 
Tables II and III. 

' A. W. Bosworth and L. L. Van Slyke: this Journal, xix, p. 67, 1914. 



142 Condition of Casein and Salts in Milk 



The data in Table II afford a basis for ascertaining the quantita- 
tive relation between casein and the phosphates. If casein is 
chemically combined with phosphates hi milk, there should be a 
fairly definite and uniform relation between these constituents 
in the insoluble portion of milk, or, stated in another way, the 
organic phosphorus of casein should show a somewhat uniform 

tabLe II. 
Amounts of proteins, casein, and phosphorus in milk. 



cow NO. 



ii) 



I BTAOE 
,OF LAC- 
TATION 



TOTAL 
PRO- 
TEIN 



PHOSPHORUS 



Insoluble 



CASEIN 



I 



ToUl Soluble 



1 
2 
3 
3 
4 
5 
5 
6 
7 
t 

8 
9 
10 
II 
12 
13 



(2) I (3) 
\per cent 

3dy8.| 4.35 



(4) 



(5) 



Total 



'Clonic 



(6) 



(7) 



Inor- 

I 
I 



(8) 



phat«9)! 

(9) 



RATIO or 

ORGAXIC TO 
INBCJLUBLE 
INORGANIC 

PHOSPHORUS 



Organic P: 
inorganic P 



(10) 



1 mo. 
1 mo. 
11 mo. 
3 mo. 
3 mo. 
7 mo. 



3.31 
3.53 
4.91 
3.93 
3.45 
3.45 



5 mo.' 4.05 
6mo.! 4.07 
10 mo.' 4.80 
7 mo.' 4.39 
8 mo. 4.33 
9 mo. 



10 mo. 

11 mo. 

12 mo. 



3.65 
4.17 
4.35 
5.71 



percent 

3.48 
2.73 
2.78 
4.09 
3.09 
2.88 
2.70 
2.92 
3.40 
3.56 
3.58 
3.47 
3.10 
3.36 
3.14 
4.97 



\per cent Txr cent per cent per cent per cent 

'o . 1 2720 . 0818 . 045^0 . 0247 . 0207 
0.11500. 05950.05550. 01940. 0361 
O.iC; no 04940. 051610. 0197 0.0319 
0. 11110. a5360. 05750.02900. 0285 
. 12780 .0563J0 .07150 .0219J0 .0496 
.09430 .04750 .04680 .0204|o .0264 
0.08700.03340.053^0.01920.0344 
.10080 .03560 .0651^0 .0207;'o .0445 
. 10630 .054a:0 .051^0 .024l[0 .0274 
0.10100.03400.06700.0253:0.0417 
0.11570.05500.06070.02540.0353 
. 10360 .0364;0 .06720 .02460 .0426 
. 1097 . 0610 . 04870 . 0220J0 . 0267 
. 10900 .04340 .065do .023910 .0417 
. 10600 .02860 .077^0 .0223{0 .0551 
. 13100 .0442,0 .0868i0 .035a!o .0515 



0868|( 



1:0.83 
1: 1.86 
1: 1.62 
1:0.98 
1:2. -26 
1: 1.29 
1 : 1 .79 
1:2.15 
1:1.14 
1: 1.65 
1 : 1 .39 
1 : 1 .73 
1:1.22 
1:1.74 
1:2.47 
1:1.46 



ratio to the insoluble inorganic or phosphate phosphorus. In 
column 10 of Table II we give the results of calculations based 
on our data, which show the amount of insoluble inorganic phos- 
phorus for one part of organic (casein) phosphorus. It is seen 
that the ratio varies between the wide limits of 1:0.83 and 
1:2.47. Even in the case of milk from the same animal at differ- 
ent stages of lactation, the proportional amounts of inorganic 
phosphorus vary widely, as from 0.98 to 1.62 with Cow 3, from 



L. L. Van Slyke and A. W. Bosworth 143 

1.29 to 1.79 with Cow 5, and from 1.14 to 1.65 with Cow 7. 
The only conclusion furnished by these results is that there is 
no evidence of chemical combination between the casein and 
the phosphates of milk. Additional evidence in confirmation of 
the foregoing statement will be furnished later in connection 
with the discussion of another phase of the subject. 

Another interesting point connected with insoluble phosphates 
and casein in milk is as to the exact compound of calcium phos- 
phate and of calcium caseinate existing in the milk. Soldner's 
inferential statement that milk casein is neutral calcium casein- 
ate (containing about 1.07 per cent of calcium), has been generally 
accepted, not so much because of positive proof but because of 
absence of any proof to the contrary. Regarding the form of the 
comp)ound in which phosphates exist in milk, all three forms 
(mono-, di-, and tri-basic phosphates) have been thought to be 
present. The insoluble phosphates have been regarded as a mix- 
ture of di- and tri-calcium phosphate. Bearing on this question, 
we present data embodied in Tables III and IV. 

The data in Table IV are derived by calculation from the figures 
given in Tables II and III, for the purpose of reducing them to a 
uniform basis that permits us to make comparison more easily. 

In our previous work we have shown that one gram of uncom- 
bined casein combines with 9 X 10~^ gram equivalents of calcium to 
form a salt that is neutral to phenolphthalein.* In column 2 of Ta- 
ble IV we use this fact in calculating the acid equivalents of the 
casein as found in each sample. In column 3 we calculate the acid 
equivalents of the insoluble inorganic phosphorus in each sample of 
milk (regarding phosphoric acid as a di-valent acid and CaHP04 
neutral to phenolphthalein). In column 4 we give the sums ob- 
tained by adding the figures in columns 2 and 3 in the case of each 
sample of milk. In columns 5 and 6 we give the combining equiv- 
alents of calcium and magnesium, and in column 7 their sums for 
each sample of milk. If now we compare in the case of each milk 
the figures contained in column 4 with those contained in column 
7, we notice that they are in close agreement, the differences being 
shown in column 8. This agreement means that the quantitative 

* A. W. Bosworth and L. L. Van Slyke: this Journal^ xiv, p. 207, 1913; 
New York Agrictdlural Experiment Station Technical IMletins, No. 26, p. 12, 
1912. 



144 Condition of Casein and Salts in Milk 



relation between the bases (calcium and magnesium) and the 
acids (casein and phosphoric acid) is that required, theoretically, 
to give di-calcium phosphate with a trace of di-magnesium phos- 
phate and the calcium caseinate neutral to phenolphthalein, in 
which casein is combined with eight equivalents of calcium 
(casein Ca^). However, the same analytical figures can with equal 
correctness be interpreted to prove that the compounds are pres- 
ent as acid caseinate and tri-calcium phosphate. 

TABLE III. 
A7nount8 of calcium and magnesium in insoluble portion of milk. 



cow NO. 



1. 
2. 
3. 



STA.OK OF 
LACTATION 



3 dys. 
Imo. 
Imo. 

3 1 11 mo. 

4 

o 

5 

6 

7 



3 mo. 
3 mo. 
7 mo. 
t5mo. 
6 mo. 
7 1 10 mo. 



8. 

9. 
10. 
11. 
12. 
13. 



Total 

(11) 



/ mo. ! 

8 mo. 

9 mo. 

10 mo. 

11 mo. 

12 mo. 



per cent 

0.1607 
0.1381 
0.1362 
0.1559 
0.1483 
0.1396 
0.1256 
0.1413 
0.146^ 
0.1523 
0.1506 
0.1503 
0.1410 
. 1379 
0.1659 
0.2167 



CALCIUM 



Soluble I Insoluble 
(12) (13) 



per cent 

0.0734 
0.0511 
0.0544 
0.0534 
0.0343 
0.0531 
0.0454 
0.0373 
0.0526 
0.0450 
0.W39 
0.0440 
0.0543 
0.0357 
0.0414 
0.0669 



Total 

(M) 



MAGNESIUM 

Soluble 
. (15) 



per cent 

0.0873 

0.0870 

0.0818 

0.1025 i 

0.1140 J 

0.0865 

0.0802 

0.1040 

0.0938 

0.1073 

0.1062 

0.1063 

0.0867 

0.1022 

0.1245 

0.1498 



per cent 

0.0156 
0.0136 
0.0180 
0.0170 
0.0184 
0.0156 
0.0147 
0.0160 
0.0144 
0.0177 
0.0153 
0.0171 
0.0168 
0.0168 
0.0191 
0.0236 



per cent 

0.0142 
0.0117 
0.0142 
0.0156 
0.0128 
0.0124 
0.0134 
0.0127 
0.0121 
0.0127 
0.0118 
0.0126 
0.0141 
0.0119 
0.0123 
0.0163 



Insoluble 

(16) 
per cent 

0.0013 
0.0019 
0.0038 
0.0014 
0.0056 
0.0032 
0.0013 
0.0033 
0.00-23 
0.0050 
0.0035 
0.0045 
0.0027 
0.0049 
0.0068 
0.0073 






In order to decide which of these sets of compounds is present 
in milk, we have tried to make a separation of the casein and 
insoluble phosphates. The results, it will be remembered, are 
obtained by difference, the milk and serum being analyzed and 
the composition of the insoluble portion being determined by sub- 
tracting the latt<»r results from the former. It seemed desirable 
to separate milk in large amounts so as to obtain the insoluble 
portion in quantity sufficient to purify and analyze. This was 
done in the following maimer, several experiments being made. 



L. L. Van Slyke and A. W. Bosworth 145 

In the first experiment 400 pounds of milk were run through a. 
cream centrifugal separator eighteen times and the deposit ("sepa- 
rator slime") collecting on the walls of the bowl was removed 
after the 1st, 6th, 12th, and 18th runs. Each of these deposits 
was placed in a mortar and triturated with small amounts of 95 
per cent alcohol with the gradual addition of more alcohol. A 
point is reached at which the whole mass becomes jelly-like, 
after which the addition of more alcohol causes the formation of a 



TABLE IV. 
eidi and bates expreaged a 




. » 3X10"' I 13 ixiir* 

-, M.ixlirt| W.IXHT' 
- 15.0X10-' 20.1X10-1 
.■ M.BXUT'I IS-IXIO"' 

. ?7 BxKT* I sa-oxnr* 

. li.8XKr< 10.8X10-' 
. M.SXKT' 22.iXUr' 
■ ■ M,JXl(r< ■ M.TXKT' 

K.ixifl-* 

M.flXllT' 
Sa 3X10"' 
37.1X10-' 
ITOXKT' 



. IDexHT" 
,' M.OXICT' 

. w.sxio-' 
. II ixur" 
.■ fT exim 
. M axitr* 

.' M.TXIO-* 



XIO-' 



156 



u.tx 

U(X 
MIX 

M.IX 

a IX 
utx 

tttX 
H.tX 
MtX 
MIX 
M.tX 

UIX 

n.tx 



H 

ill 



I'M 



ii oxiiri 

13.2X10-' 
10 IXWT' 
iSOXlO"' 



4xio"> , !,3x; 



m 

I.TXIO^ 
6.1XID-' 
l,»X!0-i 

i.ixiin 

ISXIO" 

i.»xio-i 

1 1X10- 

1.7XI0-' 

I »xio-i 
<iir' 



fine fiocculent precipitate. (Care must be taken not to add the 
alcohol too rapidly, because then there is apt to be formed a tough, 
leathery mass, which can not be handled.) The precipitate is 
allowed to settle and, after decanting the supernatant liquid, is 
triturated with several successive portions of 95 per cent alcohol, 
99 per cent alcohol, and finally ether. It is then dried at 60°C. 
for a few hours, after which the drying is completed in a vacuum 
over sulphuric acid. The analytical results are given in Table V. 



146 Condition of Casein and Salts in Milk 



TABLE V. 



Composition of insoluble portion {** separator slinie*') of milk. 



SKUVUm OF DEPOSIT TAKKN 



After Ist run. . 
After 6th run. 
After 12th run 
After 18th run 



per cent 

86.31 
90.07 
90.84 
91.98 



M 

O 

e 

8 



3 



t 



percent 

10.43 
9.35 
9.53 
9.62 



percent 



2.011 

2.023 



s 

Jl 

per cent 






o 

|8 



g 



RATIO or 

ORGANIC TO 

INSOLUBLE 

INORGANIC 

PHOB- 

PHOBUA 



Organic P: 
I inoivanic P 



2.182 0.621 
1.950 0.649 



0.645 



per cen^, per cent 

1.56l' 3.386' 
1.301 3.246; 
1.366 3.34a 



0.662 1.361! 3 



.2231 



1:2.51 
1:2.00 
1:2.11 
1:2.06 



The figures in Table V, obtained by direct analysis of the insol- 
uble deposit or "separator slime," show a striking agreement with 
results obtained by the indirect method, which is brought out 
more clearly by expressing the above figures in the form of gram 
equivalents, as follows: 

TABLE VI. 
Amounts of acids and bases expressed as gram equivalents. 



SAMPLE OP DEPOSIT TAKEN 

After 1st run 

After 6th run 

After 12th run 

After 18th run 



CASEIN AS ORAM 

BQUIVALENTB 

or AQD 



77.7 X 10-» 
81.1 X 10-» 

80.8 X 10-* 
82.8 X 10-» 



PHOSPHATES 

AS ORAM 

EQUITALBKTS 

OP DI-BABIO 

AdD 



100.7 X 10-»178 
82.9 X 10-»164 
88.1 X.10-> 
87.8 X 10-»|l70 



SUM OP GRAM 

XQUIYALEMTS 

OP CASEIN AND 

PHOSPHATES 



4X 10-' 

X 10-^ 

168.9 X 10-^ 

6 X 10-: 



ORAM 

EQUIVALENTS 
OP CALCIUM 



169.8 X 10-* 
162.8 X 10-* 
167.2 X 10-» 
161.2 X 10-» 



The high percentage of inorganic phosphorus in the deposit 
from the first run indicates that the phosphates are heavier than 
the caseinates and could be separated from them if a certain speed 
were used in running the separator. This point is further shown 
by the following experiments. 

In the first experiment the bowl of a cream-separator was filled with f at- 
free milk (about 1000 cc.) and was whirled for two hours, at a speed of 5000 
revolutions per minute, when the milk was taken out and the ''separator 
slime'' which had collected on the bowl was removed and treated with 
alcohol and ether in the manner already described. The same milk was 
returned to the separator bowl and again whirled for two hours, when the 



L. L. Van Slyke and A. W. Bosworth 147 



deposit was again removed and treated as before. When removed the sec- 
ond time, that is, after four hours of whirling, the milk was nearly as clear 
as whey, most of the suspended phosphates and casein having been deposited 
on the walls of the bowl during the whirling. The results of analysis of 
the "separator slime" deposited after each two hours' whirling are given in 
Table VII. 

These results show that two-thirds of the insoluble inorganic 
phosphorus was removed during the first two hours of whirling, 
again indicating that the phosphates are heavier than the casein. 
The ratio of casein to phosphates is here also shown to be wholly 
irregular, indicating no definite combination. 

Expressing the data in Table VII in the form of gram equiv- 
alents, we have the figures contained in Table VIII. 

TABLE vu. 
Composition of insolvble portion of milk deposited at different intervals. 



"8LIMB" 

PORMBD 

BT 

WHIRUNO 

pob2sb. 

PERIODS 



l8t2hrs. 
2d2hr8. 



CASBIN 



per cent 

90.68 
91.12 



ASH 



TOTAL 

PHO»- 

PHOBUB 



per cent 

9.32 
8.88 



per cent 

1.909 
1.437 



PHOB- 

PHOBUB IN 

CASBIN 



per cent 

0.653 
0.656 



PHOB- 
PBOBUB AS 
PHOS- 
PHATES 



per cent 

1.256 
0.781 



CALCIUM 


RATIO OP 
ORGANIC TO 
INIOLUBLB 
INORGANIC 
PHOS- 
PHORUS 




inorganio P 


per cent 

3.090 
2.601 


1:1.92 
1: 1.19 



An examination of these figures shows that there is the same 
balance between the acids (casein and phosphoric acid) and the 
bases (calcium and magnesium) in the two separate deposits, 
even when the inorganic phosphorus is so unevenly distributed 
l)etween them, which furnishes proof for two points: (1) The 
inorganic phosphorus must be in the form of neutral calcium 
phosphate (CaHPOO; for otherwise the balance between bases 
and acids would be altered, acid calcium phosphate (CaH4P208) 
giving an excess of acid, and tri-calcium phosphate (CajP208) an 
excess of base in the "slime" deposited in the first whirling. (2) 
If the phosphates were in combination with the casein we should 
expect to find the ratio between the organic phosphorus and the 
inorganic phosphorus the same in both deposits, but instead of 
uniformity wjp find the ratio showing as wide a variaticfn as 1:1.92 
and 1 : 1.19 in the two cases. 



148 



Condition of Casein and Salts in Milk 



In the second experiment further evidence is furnished, showing 
that neutral calcium phosphate (CaHP04) is a normal constituent 
of milk. 

Four 500 cc. bottles were filled with separator skim-milk to which some 
formaldehyde had been added, and after standing at room temperature for 
four days, were whirled in a Bausch and Lomb precision centrifugal ma- 
chine for thirty minutes at a speed of 1200 revolutions per minute. A sedi- 
ment was deposited, which after purification by treatment with alcohol 
and ether, as previously described, weighed 0.4 gm. 

Analysis of this gave the following results: casein, 20.78 per 
cent; total phosphorus, 18.38 per cent; phosphorus combined with 
casein, 0.15 per cent; phosphorus combined as phosphates, 18.23 
per cent; calcium, 22.79 per cent; ratio of organic phosphorus to 
inorganic phosphorus, 1: 121; casein as gram equivalents of acid, 

TABLE viu. 
Amounts of acids and bases expressed as gram equivalents. 



CASEIN AB GBAM 
EQUIVALBNT8 

I OP ACm 



PHOSPHATES 
AS GBAM 

EQUIVALENTB 

OP OI-BASIC 

ACID 



SUM OP GRAM 

EQUIVALENTS 

OP CASEIN AND 

PHOSPHATES 



1st deposit ; 81 .6 X 10-» 

2d deposit 82.0 X 10-» 



81.1 X 10-»ie2.7 X 1M164.6 X lO"' 
50.4 X 10-»182.4 X ia-«184.6 X 10-» 



GRAM 

EQUIVALENTS 

OP CALCIUM 



18.7 X 10~^; phosphates as gram equivalents of di-basic acid, 
1175 X 10"^; sum of casein and phosphates as gram equivalents 
of acid, 1194 X 10""'; gram equivalents of calcium, 1140 X 10""'. 

In these figures we again find the same balance between bases 
and acids, which can mean only that the phosphate compound 
deposited is di-calcium phosphate (CaHPOO. The degree of 
centrifugal force developed was sufficient to throw out a relatively 
large amount of di-calcium phosphate, but not powerful enough 
to throw out very much casein, thus serving as a means of efifecting 
a nearly complete separation of these two constituents. 

Babcock^ whirled skim-milk in a separator for several hours, 
removing portions from time to time for analysis and finally 
determined the amounts of casein,' calcium, and phosphorus in 

' S. M. Babcock: Wisconsin Agricultural Experiment Stationy 12th An-- 
nual Report, p. 93, 1895. 



L, L. Van Slyke and A. W. Bosworth 149 

the deposited "slime." While the experiments were preliminary 
in character and the results not sufficient to base permanent con- 
clusions on, they tended to show that the casein and phosphates 
were not in combination. From the analjrtical results showing 
the relation of calcium to phosphorus, the conclusion was drawn 
that tri-calcium phosphate is the compound present in milk. The 
figures for calcium and phosphorus were based upon the total 
amounts contained in the deposit and no allowance was made for 
the calcium in combination with casein and the phosphorus of the 
casein. This fact accounts for the difference between the results 
reported by him and the conclusions reached by us. A recal- 
culation of his data, after deducting the amounts of calcium and 
phosphorus combined with casein, gives figures that correspond 
to the composition of CaHPO* and not Ca8P208, thus confirming 
the results of our work. 

Acidity of milk and milk serum. 

Both fresh milk and the serum from fresh milk show a slight 
acid reaction to phenolphthalein. This has been believed to be 
due to casein or acid phosphates in the milk or to both. The fact 
that fresh milk and its serum are strongly alkaline to methyl 
orange indicates that the acidity is due to acid phosphates, though 
it does not necessarily show that acid caseinates are not also 
responsible for some of the acidity. The results of our work given 
in the preceding pages furnish aid in determining to what com-» 
pounds in milk the acid reaction to phenolphthalein is due. 

A 1000 cc. sample of milk was obtained from each of eight cows 
immediately after milking, and chloroform (50 cc.) was added to 
this at once. The acidity of the milk and of the serum was de- 
termined after treatment with neutral potassium oxalate accord- 
ing to the method of Van Slyke and Bosworth." The results are 
given in Table IX. 

These figures show that the acidity of fresh milk is the same 
as that of its serum, which means that the constituents of the 
milk causing acidity are soluble constituents contained in the 
serum. Since the serum contains phosphates in amounts suffi- 
cient to furnish two to four times the acid phosphates required 

•L. L. Van Slyke and A. W. Bosworth: this Journal, xix, p. 73, 1914. 



150 Condition of Casein and Salts in Milk 

to account for the acidity, and since, moreover, no other acid 
constituents of the milk sermn are present in more than minute 
quantities, and are wholly insufficient to cause the observed 
degree of acidity, it appears a reasonable conclusion that the 
acidity of fresh milk is due to soluble acid phosphates. This 
conclusion is further strengthened by the results given in the 
preceding pages which go to show conclusively that the insoluble 
constituents of fresh milk are neutral in reaction, consisting largely 
or wholly of neutral calcium caseinate (casein Ca^), neutral di- 
calcium phosphate (CaHPQO, and fat. 



TABLE IX. 
Acidity of milk and milk serum. 



N 



NO. OF BAMPLS 



I 



NO. or OC. OP jjf ALKALI BEQUIBED TO 
NEUTBAUZB 100 OC. OP 



1 


MUk 

4.8 
6.2 
4.2 
6.0 
6.4 
4.4 
7.0 
6.6 


Milk serum 
5.0 


2 


6.2 


3 


4.2 


4 


5.8 


5 


6.4 


6 

7 


4.4 
6.8 


8 


6.4 







Compounds of milk. 

It is difficult to learn what are the individual forms or compounds 
hi which the salts exist in milk. Attempts have been made to 
determine this by inference based on analytical results. In view 
of the data presented in the preceding pages, taken together with 
many other analytical data worked out by us, we suggest the fol- 
lowing statement as representing in some respects more closely 
than pre\'ious ones the facts corresponding to our present knowl- 
edge of the prmcipal constituents of milk. The amounts are 
based on milk of average composition.^ 



^L. L. Van iSlykc: Modern Methods of Testing Milk and Milk Products, 
Xcw York, 1012.* 



L. L. Van Slyke and A. W. Bosworth 151 

Percent 
Fat 3 . 90 

Milk sugar 4.90 

Proteins combined with calcium 3 .20 

Di-calcium phosphate (CaHP04) 0.175 

Calcium chloride (CaCU) 0.119 

Mono-magnesium phosphate (MgH^PjOg) . 103 

Sodium citrate (NajCeHjOT) 0.222 

Potassium citrate (KiCeHjOy) 0.052 

Di-potassium phosphate (K1HPO4) 0.230 

Totalsolids 12.901 

CONCLUSIONS. 

1. Milk contains two general classes of compounds, those in 
true solution and those in suspension, or insoluble. These two 
portions can be separated for study by filtering the milk through 
a porous earthenware filter like the Pasteur-Chamberland filtering 
tube. 

, 2. Serum prepared from fresh milk is yellow with a faint green- 
ish tinge and slight opalescence. The following constituents of 
milk are wholly in solution in the milk serum: sugar, citric acid, 
potassium, sodium, and chlorine. The following are partly in 
solution and partly in suspension : albumin, inorganic phosphates, 
calcium, magnesium. Albumin in fresh milk appears to be ad- 
sorbed to a considerable extent by casein and therefore ovAy a 
part of it appears in the serum. In serum from sour milk and milk 
to which formaldehyde has been added, nearly all of the albumin 
appears in the serum. 

3. The insoluble portion of milk separated by filtration through 
Pasteur-Chamberland filtering tubes is grayish to greenish white 
in color, of a glistening, slime-like appearance, and of gelatinous 
consistency. When shaken with water it goes into suspension, 
forming a mixture having the opaque, white appearance of milk. 
Such a suspension is neutral to phenolphthalein. When purified, 
the insoluble portion consists of neutral calcium caseinate (casein 
Ca4) and neutral di-calcium phosphate (CaHP04). The casein and 
di-calcium phosphate are not in combination, as shown by a study 
of sixteen samples of milk from thirteen indi\'idual cows, and also 
by a study of the deposit or "separator slime" fonned by whirl- 
ing milk in a cream-separator. By treating fresh milk with for- 



152 Condition of Casein and Salts in Milk 

maldehyde and whirling in a centrifugal machine under specified 
conditions, it is possible to effect a nearly complete separation of 
phosphates from casein. 

4i^oth fresh milk and the serum from fresh milk show a slight 
acid reaction to phenolphthalein, but are strongly alkaUne to 
methyl orange, indicating that acidity is due, in part at least, 
to acid phosphates. In eight samples of fresh milk the acidity 
of the milk and of the milk serum was determined after treatment 
with neutral potassium oxalate. The results show that the acid- 
ity of the whole milk is the same as that of the serum, and that, 
therefore, the constituents of the serum are responsible for the 
acidity of milk. There is every reason to believe that the phos- 
phates of the serum cause the observed acidity. 

5. The data presented, taken together with results of other work, 
furnish a basis for suggesting an arrangement of the individual 
compounds contained in milk, especially including the salts. ' 



RESEARCHES ON PURINES. XVI.> 

ON THE ISOMERIC MONOMETHYL DERIVATIVES OF 2-METHYL- 

MERCAPTO-4-AMINO-6-OXYPYRIMIDINE. ON 1-METHYL- 

2-METHYLMERCAPT0.6,^DIOXYPURINE. 

By carl O. JOHNS and BYRON M. HENDRIX. 
{From the Sheffield Laboratory of Yale University, New Haven.) 

(Received for publication, December 29, 1914.) 

When 2-thio-4-amino-6-oxyp3rrimidine (I)^ is alkylated by means 
of methyl iodide or dimethyl sulphate, position 2, which contains 
the sulphur atom, is the first to be attacked, and 2-methylmer- 
capto-4-amino-6-oxypyrimidine (III)' is formed. That the methyl 
group has become attached to sulphur is shown by the fact that 
methylmercaptan is evolved when the pyrimidine is boiled with 
concentrated hydrochloric acid. 

When 2-methylmercapto-4-amino-6-oxypyrimidine (III) was 
methylated with dimethyl sulphate, a second methyl group was 
attached to the pyrimidine ring and two isomeric methyl deriva- 
tives were formed. One of these was produced only in small 
quantity, about 20 per cent of the calculated yield. This com- 
pound was very soluble in ether and melted at 144°C. Its isomer, 
which was the chief product of the reaction, was not soluble in 
ether and melted at 255**C. 

It seemed probable that 2-methylmercapto-4-amino-6-oxypy- 
rimidine (III) would methylate on the nitrogen atom in position 
1 or on the oxygen atom in position 6, forming either 1-methyl- 
2-methylmercapto-4-amino-6-oxypyrimidine (VI) or 2-methyl- 
mercapto-4-amino-6-methoxy pyrimidine (IV). The compound 
which melted at 144°C. and which was soluble in ether proved 
to be the methoxy' derivative (IV). This was shown by pre- 

*C. O. Jbhns and B. M. Hendrix: this Journal, xix, p. 25, 19U. The 
present investigation was aided by a grant from the Bache fund. 

* VV. Traube: Ann, d, Chem., cccxxxi, p. 71, 1904. 

'Johnson and Johns: Am, Chem. Jour., xxxiv, p. 181, 1905; C. O. Johns 
and E. J. Baumann: this Journal, xiv, p. 384, 1913. 

15.3 



154 Researches on Purines 

paring it by the action of sodium methoxide on 2-methylnier- 
capto-4-amino-6-chlorpyrimidine (V)/ These substances reacts! 
smoothly to give 2-methylmercapto-4-amino-6-methoxypyrimidine 
(IV), which was found to be identical with the compound which 
melted at 144*^C. 

It is, therefore, probable that the compound which melted at 
255°(). was l-methyl-2-methylmercapto-4-amino-6-oxypyrimidine 
(VI). This view is supported by the fact that in the case of all 
the pyrimidines which we have methylated and which contained 
an amino group in positions 4 or 6, the methyl group has attached 
itself to the nitrogen atom farthest away from the amino group. 
Thus, pyrimidines containing an amino group in position 6 have 
methylated in position 3, or what might be termed the position 
"para" to the amino group.* If this rule holds, 2-methylmer- 
capto-4-amino-6-oxypyrimidine (III) would methylate in position 
1, and the methyl derivative would be l-methyl-2-methylmercapto- 
4-amino-6-oxypyrimidine (VI). This compound was obtained in 
this lal)oratory some years ago as a by-product in methylating 
2-thio-4-amino-6-oxypyrimidine (I) in the presence of an excess 
of sodium ethylate and methyl iodide. We supposed at that 
tune that it was a methoxy derivative and suggested that its 
fornmla was 2-methylmercapto-2-amino-6-methoxypyrimidine 
(IV).' The results of the present investigation show that we 
were mistaken, because the compound melted at 255°C. and 
was identical with l-methyl-2-methylmercapto-4-amino-6-oxypy- 
rimidine (VI). 

Sin(^c two isomeric methyl derivatives are formed when 2-methyl- 
mercapt()-4-amino-6-oxypyrimidine (III) is methylated, it is prob- 
able that this compound exists in two tautomeric forms. These 
two forms may be represented by formulas II and III. On 
niothylat ion the fonner would give a 6-methoxy derivative and the 
latter a 1 -methyl compound. The supposition that it exists 
in the cnol form is supported by the fact that it gives 2-methyl- 
mercapto-4-ainino-6-chlorp5rrimidine (V)^ when boiled with phos- 

* Johnson and Johns: Am. Chcjn. Jour., xxxiv, p. 183, 1905. 

* Johns: this Journal, xi, p. 75, 1912; Johns and Baumann: ibid., xvi, 
!>. 137. 1913. 

•Johnson and Johns: Am. Chrm. Jour., xxxiv, p. 182, 1905. 
'Johnson and Johns: loc. cit. 



C. O. Johns and B. M. Hendrix 



155 



phorus oxychloride, a reaction in which an OH group must take 
part. It is also noteworthy that the yield of the chloride is less 
than 50 per cent of the calculated quantity even when a very 
great excess of phosphorus oxychloride is used, and that a large 
portion of the 2-methylmercapto-4-amino-6-oxypyrimidine (III) is 
recovered unaltered. 

When 1-methyl - 2 - methylmercapto - 4 - amino - 6 - oxypyrimidine 
(VI) was treated with nitrous acid, an almost quantitative yield 
of l-methyl-2-methylmercapto-4-amino-5-nitroso-6-oxypyrimidine 
(IX) was obtained. By means of ammonium sulphide this com- 
pound was reduced to l-methyl-2-methylmercapto-4,5-diamino- 
6-oxypyrimidine (VIII). When the latter compound was heated 
with urea, an almost quantitative yield of l-methyl-2-methyl- 
mercapto-6,8-dioxypurine (VII) was obtained. 

These researches will be continued. 



KS--XCO 



SC CH 



HN— CNH, 
I 



N=COCH, 



CH^SC CH 

II 



N— CNH, 
IV 



CH,N— CO 



CH,SC C— NH < — 

>C() 
-NH 



VII 



N=COH 



HX— CO 



^ CH,SC CH 



N— CNH2 
N=rCCl 



CH,SC CH 



N— CNH, 
V 



CH,N— CO 



CHaSC CNH, ^ 



N— CNH, 
VIII 



or 



CHaSC CH 



N— CXH, 
III 



CHaX— CO 
CHaSC CH 



X— CXH, 
VI 



CH,X— CO 
CH,SC CXO 



N— CXH, 
IX 



156 Researches on Purines 

EXPERIMENTAL PART. 

Methylation of 2'methylmer<xiptO'4-^min(h6-oxypynmidine, 

Ten grams of 2-methylinercapto-4-aniino-6-oxypyrimidinc^ were 
dissolved in 65 cc. of normal potassium hydroxide solution, and 
9 grams of dimethyl sulphate were added gradually while the 
solution was agitated by frequent shaking. A white, crystalline 
precipitate began to appear almost inmaediately, and this soon 
became very bulky. As soon as the solution became acid to 
litmus the crystals were filtered off by suction. The filtrate was 
neutralized with sodium hydroxide and evaporated to dryness. 
The residue was washed with cold water, the solid was filtered 
off and added to the crystals abeady obtained. The combined 
solids were then triturated with dilute ammonia to dissolve any 
unaltered 2-methylmercapto-4-amino-6-oxypyrimidine, a small 
quantity of which was found to be present. That part of the 
residue which was not soluble in ammonia consisted of two com- 
pounds which differed widely as to their melting points and solu- 
biUty in ether. The compound having the lower melting point 
was very soluble in ether, while the one with the higher melting 
point was almost insoluble in this solvent. Ether, therefore, 
served as a means of separating these compounds from each other. 

2-Methylmeraipto-4'^mino-6'methoxypyrimidine. 

N=COCH, 

I I 
CHiSC CH 

II II 
N— CNH, 

The ether extract which was obtained in the above experiment 
was evaporated to dryness. A white crystalline residue remained. 
This was recrystallized from dilute alcohol by dissolving it in 
about 50 per cent alcohol and concentrating the solution until 
crystals appeared on cooling. In this manner a beautiful, crystal- 
line substance was obtained. This was very soluble in alcohol, 
ether, or benzene. It was moderately soluble in hot and difficultly 
soluble in cold water. The crystals sublimed slowly when heated 

■Johns and Baumann: loc. cit. 



C. O. Johns and B. M. Hendrix 157 

to about 90®C. Both the crystals and their sublimate melted 
to an oil at 144^0. The yield was 20 per cent of the calculated 
quantity. 

Calculated for 
CcHtONsS: Found: 

N 24.57 24.69 

The structure of the compound just described was ascertained 
in the following manner: 

5.8 grams of 2-methylmercapto-4-amino-6-chlorpyrimidine* were 
dissolved in 50 cc. of methyl alcohol to which 1.6 grams of metallic 
sodium had previously been added. This mixture was heated 
under a reflux condenser on a water bath for an hour. Sodium 
chloride was deposited during the process of heating. The alco- 
hol was removed by evaporation on a steam bath. The residue 
was stirred with water and this mixture was acidified with hydro- 
chloric acid. This treatment left a white residue which was 
recrystallized from dilute alcohol. The crystals melted at J.43°C. 
and had all of the properties of the compound which we had 
obtained from the ether extract, as described above. 

Calculated for 
C«HffONsS: Found: 

N 24.57 24.38 

Samples of the compound obtained from the ether extract and 
by the action of sodium methoxide on 2-methylmercapto-4-amino- 
6-chlorpyrimidine were mixed and the mixture was found to melt 
at 144°C. Hence the two were identical. 

l'MethyU2-m£thylmer(xipto-4-^min(h6^xypyrimidine, 

CH,N— CO 



CHiSC CH 

!! II 

N— CNH, 

This was the chief product of the reaction between dimethyl 
sulphate and 2-methylmercapto-4-amino-6-oxypyrimidine, under 
the conditions described above, and was the portion not soluble 
in ether. After dissolving out the 2-methylmercapto-4-amino-6- 

' Johnson and Johns: loc. cii. 



158 Researches on Purines 

inethoxypjrimidine by means of ether and filtermg, the solid 
residue was recrystallized from alcohol. It separated in the form 
of slender prisms. These melted to an oil at 255°C. They were 
moderately soluble in hot water or alcohol, but almost insoluble 
in ether or benzene. The yield was 60 per cent of the calculated 
quantity. 

Calculated for 
C«HffONsS: Found: 

X 24.57 24.71 

This substance agreed in all respects with the compound which 
had previously been obtained in this laboratory as a by-product 
by the action of an excess of methyliodide and sodium ethylate 
on 2-thio-4-amino-6-oxypyrimidine. We assumed at that time 
that this compound was 2-methylmercapto-4-amino-6-methoxy- 
pyrimidine. The results of the present investigation indicate 
that it was l-methyl-2-methylmercapto-4-amino-6-oxypyrimidine. 

l^yfethyl'^^nethylmeraiptO'i-^mino-O'nitroso^'Oxypyrimidm^ 

CH,N— CO 



CHjSC CNO 

II II 
N— CNH, 

4.8 grams of l-methyl-2-methylmercapto-4-amino-6-oxypyrimi- 
dinc were suspended in an excess of dilute hydrochloric acid and 
a solution of 2.5 grams of sodium nitrite in a little water was 
added gradually. The suspended material assumed a deep blue 
color and became crystalline and very bulky. The reaction was 
complete in about fifteen miimtes, after which the soUd material 
was filtered off and washed with cold water. The portion used 
for analysis was recrystallized from water and was so obtained 
in the form of slender prisms of a deep blue color. These melted 
at 235°C. They were sparingly soluble in hot water or alcohol 
and almost insoluble in benzene or ether. The aqueous and alco- 
holic solutions possessed the deep blue color of an ammoniacal 
copper sulphate solution. The yield was abnost quantitative. 

Calculated for 
C<H«OjN4S: Found : 

N 28.00 27.94 



C. O. Johns and B. M. Hendrix 159 

UMeihyU2'^m£ihylmercapU)^4j6'diamino^ 

CH.N— CO 



CHiSC CNH, 

II II 
N— CNH, 

5.2 grams of l-methyl-2-methylmercapto-4-ainino-6-iiitroso-6- 
oxyp3rrimidine were finely pulverized and suspended in 150 cc. 
of water. The nitroso compound was reduced by adding ammon- 
ium sulphide gradually until an excess seemed to be present. 
The reduction proceeded very rapidly and the nitroso compound 
all dissolved, sulphur being precipitated. After a half hour the 
reaction seemed complete and the sulphur was filtered off. The 
filtrate was concentrated to about 50 cc. to remove more sulphur 
and drive off the excess of ammonium sulphide. On filtering 
again and cooling, a crystalline precipitate was obtained. This 
was recrystallized from water and it separated in the form of 
aeicular prisms, which melted at 212°C. They were moderately 
soluble in hot water or alcohol, but almost insoluble in ether or 
benzene. The yield was only about 50 per cent of the theoretical 
quantity. 

Calculated for 
C«Hi-0N«S: Found: 

X 30.11 30.06 

1 'Methyl'g-melhylmercapUhGyS'dioxypunne. 

CH,N— CO 



CHiSC C— NH 

II II >co 

N— C— NH 

1-Methyl -2-methylmercapto -4,5- diamino - 6- oxy pyrimidine was 
heated with an equal weight of urea in an oil bath at 160-170°C. 
for an hour. The mixture melted and amimonia was evolved, 
then it solidified to a hard cake. This material was pulverized 
and washed with cold water. The residue was dissolved in dilute 
anunonia to which a little sodium hydroxide was added. On fil- 
tering and acidifying with acetic acid, a powder was precipitated. 



i6o Researches on Purines 

This was composed of extremely small crystals. These decomposed 
above 300°C. without melting. They were slightly soluble in hot 
water or alcohol but did not dissolve in ether or benzene. 
They did not give the murexide reaction. The yield was almost 
quantitative. 

Calculated for 
C7HsOiN4S: Found: 

X 26.41 26.46 



LlPmS IN NUTRITION. 

By C. G. MacARTHUR and C. L. LUCKETT. 
(From the Department of Biological Chemistry , University of Illinois ^ Urbana. ) 

(Received for publication, December 19, 1914.) 

There is considerable diflference of opinion as to the necessity 
of certain lipin-like substances for maintenance and growth. The 
earlier investigators did not consider the lipins as an essential 
part of a food. But recently several important papers^ have 
appeared indicating the essential presence in complete rations 
of a substance with fat-like solubilities. There is a rather wide 
range of opinion as to the nature of this material. An examina- 
tion of the articles just referred to will show that the subject has 
been approached from several angles; and though the fats, choles- 
terol, the phosphatids, and the cerebrosides have at one time or 
another been suspected of being the essential substance, they 
are not now generally supposed to be so. 

One of two objections can be offered to most of the previous 
investigations on this subject. In some the food used did not 
consist of pure, definite substances; all the difficulties encountered 
in an unknown mixture apply here. In others the synthetic food 
was probably not lipoid-free. Ether does not remove all the 
Upoids; even cold alcohol fails to extract them completely; hot 
alcohol or a hot benzene and alcohol mixture seems to be necessary. 

This investigation was carried out with the hope that by mak- 

* W. Stepp: Biochem, Ztschr., xxii, p. 452, 1909; Ztschr.f, Biol., Ivii, p. 
• 135, 1912; Ixii, p. 405, 1913. 

L. B. Mendel and T. B. Osborne: this Journal^ xii, p. 81, 1912; xiii, p. 
233, 1912-13; xvi, p. 423, 1913-14. 

E. V. McCoUum, J. G. Halpin, and A. H. Drescher: this Journal^ xiii, 
p. 219, 1912-13. 

E. V. McCollum andM. Davis: this Journal, xv, p. 167, 1913. 

C. Funk: Jour. Physiol., xliii, p. 395, 1911-12; xliv, p. 50, 1912. 

E. A. Cooper: Biochem. Jour., viii, p. 347, 1914. 

F. G. Hopkins: Jour. Physiol., xliv, p. 425, 1912. 

i6i 
rfu lomoiAL or biolooxcai. chbiostbt, vol. xx, no. 2 



1 62 Lipins in Nutrition 

ing the food used largely of definite substances and by thoroughly 
extracting the substances fed, when contamination with lipins 
was at all possible, some clearer ideas of this uncertain substance 
might be obtained. 

Preparation of food. 

The lard was of the best quality and gave no nitrogen test. 
The starch and lactose were phosphorus- and nitrogen-free. 

The ordinary casein, referred to as casein (1), was refluxed with 
alcohol for three two day periods, and was filtered and pressed 
after each period. It was finally washed with ether and dried. 
In the later experiments the purest casein (2) (Hammarsten) was 
fed. 

The part of the ration most difficult to free from lipins is the 
salts. Salt mixture^ (i) was not adequate, although used as 

SALT MIXTURE (i). 

gm. 

Calcium phosphate 10 

Potassium hydrogen phosphate 37 

Sodium chloride 20 

Sodium citrate 15 

Magnesium citrate 8 

Calcium lactate 8 

Iron citrate 2 

supplementary food. A more successful combination of salts^ 
is mixture (ii). Probably the most satisfactory purely synthetic 

SALT MIXTURE (ii). 

gm. 

Hydrochloric acid 12 .75 

Phosphoric acid 10.32 

Citric acid 10.10 

Sulphuric acid 0.92 

Calcium carbonate 13 .84 

Magnesium carbonate 2 .42 

Potassium carbonate 14. 13 

Sodium carbonate 14.04 

Iron citrate .634 

salt mixture (iii) contains these ingredients:* 

* F. Rohmann: Jahreab. f. Thier-Chem., xxxviii, p. 659, 1909. 

^ Osborne and Mendel: Ztschr. f. physiol. Chem., Ixxx, p. 307, 1912. 

* McCollum (unpublished). 



C. G. Mac Arthur and C. L. Luckett 163 

SALT MIXTURE (iii) . 

gm. 

Sodium chloride 0. 146 

Magnesium sulphate .225 

Sodium dihydrogen phosphate .293 

Potassium monohydrogen phosphate .805 

Calcium tetrahydrogen phosphate .456 

Iron lactate (Merck) . 100 

2.4 gm. of the above and 1.3 gm. of calcium lactate were used with each 
100 gm. of ration. 

In some of the experiments the salts, carbohydrate, and part 
of the protem were supplied by milk powder^ (iv), which was 
treated like the casein, with alcohol and ether, because it had 
been indicated in preliminary experiments that the milk powder 
contained the substances to be investigated. 

However, protein-free milk* (v) is the surest source of salts. 
They were prepared in the usual way, then extracted with hot 
alcohol and ether alternately, twice with each solvent. After 
the first experiments it was found that three one day periods of 
extraction with a mixture of benzene and absolute alcohol (60 
per cent benzene, 40 per cent alcohol) were more efficient in 
removing fats and less liable to remove necessary salts. After 
this treatment the benzene was removed from the residue by 
washing thoroughly with ether. Protein-free milk contains 
lactose. 

To obtain Upoids the yolks of eggs were spread on plates and 
dried in an air drier at room temperature. The dry yolk was 
extracted with ether and the egg cephalin (a) precipitated with . 
alcohol. The filtrate was evaporated carefully to dryness at low 
temperature, taken up in ether, and the egg lecithin (b) precipi- 
tated by aceton. The filtrate was slowly evaporated and labeled 
the third fraction (c). This third fraction is almost entirely ordi- 
nary fats and cholesterol. Some mice were fed on unseparated 
alcohol and ether extract (d), which consists of cephaUn, lecithin, 
small amounts of other phosphatids, fats, cholesterol, cerebro- 
side-like substances, and an unknown compound or compounds. 
In order to study more completely the alcohol extract, the dried 

^ Merrell Soule's skimmed milk powder. 

• Osborne and Mendel: Carnegie Institution of Washington Publications , 
No. 156, pt. u, 1911. 



164 Lipins in Nutrition 

yolks were in some cases treated with hot alcohol and allowed 
to cool. This gave a cold alcohoUsolvble part (e), consisting largely 
of lecithin, cholesterol, fats, some cerebrosides, and miknown 
material, and it gave also a hot alcohol-soluble part (f), consisting 
of cerebrosides, cephaUn, and some unknown substances. 

Another portion of dried egg was first extracted with cold 
alcohol. This gave the cold alcohol extract (g). In this extract 
are cephalin, lecithin, fats, cholesterol, cerebrosides, and unknown 
substances. A portion of this extract was evaporated in vacuum 
to dryness and the residue treated with ether. This gives the 
ether-solvble portion (h), which of course contains the cephalin, 
lecithin, fats, and cholesterol. The egg tissue was then treated 
with hot alcohol. This is the hot alcohol extract (i), of cerebro- 
sides, cephalm, and unknown material. 

EXPERIMENTS AND DISCUSSION. 

The white mice used in the work were kept at a rather uni- 
form temperature of 75°-80°F. in a moderately well lighted base- 
ment room. The usual precautions as to cleanliness were ob- 
served. The mice were supplied with water by the bottle hang- 
ing-drop method. They were fed regularly, usually every other 
day. The amount of food eaten was originally recorded, but 
later was only irregularly noted, because it was found that the 
amounts of food eaten indicated no points that the records of 
weight did not show, and were difficult to obtain accurately, owing 
to the scattering of the food by the mice. 

The question of palatability may enter into this work, but 
probably a food is palatable if it is complete and satisfies all 
physiological needs, while an incomplete food will not be so 
readily eaten. Whenever any food mixture was given to a mouse 
accustomed to an ordinary diet of corn, dog biscuit, sunflower- 
seeds, carrots, and meat, there was a short period of slower gro\\th, 
or even no growth, followed by nearly normal growth if the 
food mixture was complete. 

For purposes of comparison, the rate of normal growth of a 
mouse under the conditions of these experiments is given in the 
following table: 



C. G. MacArthur and C. L. Luckett 



165 





EXPE 


RIMENT 40. 






FOOD 


DATS 


WCIOHT 


Crushed com 


1 
12 
19 

26 

42 


gm. 

13.5 


Sunflower-seed 


17. 


Meat - 


21. 


Carrots 




• 


22. 


DoK biscuit - - - 


22.2 











To see if the fats themselves are indispensable, mice were fed 
lipoid-free food, to which was added olive oil, or lard, or butter. 



Experiment 14. 



FOOD 


1 

1 
1 


DATS 


WEIGHT 




gm. 




gm. 


Ether and alcohol extracted casein (1) 


18. 
22. 


1 

10 


12.7 


Ether and alcohol extracted milk powder (iv). . 


12.5 


Com starch 


23.8 
9.2 
1. 


20 
27 
34 


12.5 


Lactose 


12.9 


Salt mixture (i) 


11.5 


Lard 


25. 


41* 
48 


11.3 


•Alcohol and ether extracted egg yolk (d) added 


12.1 


on 4lRt day , , 


3. 


55 


13.3 


• 




62 


14.3 






69 


13.7 






79 


14. 


• 




83 


14. 






90 


14.8 






97 


15.6 






104 


15.7 






111 


15.9 






132 


18. 



1 66 



Lipins in Nutrition 

Experiment 15. 



WEIGHT 



POOD 



gm. 

Ether and alcohol extracted casein (1) . . . | 18. 
Ether and alcohol extracted milk powder | 

(iv) j 22. 

Com starch I 23 .8 

Lactose ■ 9.2 

2. 
25. 



DATS 



Male 



Salt mixture (i). 
Creamery butter 



10 
17 
24 
31 
38 
42 
46 



gm. 

17. 

17.9 

17.9 

15.9 

14. 

11.2 

10.2 



Female 
gm. 

15.7 

17.5 
16.2 
14.9 
13.7 
11. 
9.9 



Died. 



Experiment 16. 



FOOD 



DATS 



WEIQHT 



Male 



Ether and alcohol extracted casein (1) 

Ether and alcohol extracted milk powder (iv) 

Com starch " 

Lactose 

Salt mixture (i) 

Olive oil 



gm, 

18. 
22. 
23.8 

9.2 

1. 
25. 



1 

10 
17 
24 
31 
38 
42 



gm. 

. 14.3 

12.5 

11.3 

9.5 

8. 

7.S 

6.7 
Died. 



GMa20 



0AY5 




t40 



C. G, MacArthur and C. L. Luckett 



167 



These data strongly suggest that olive oil, lard, or butter, when 
added to the ration, do not make the food a complete one 
Jorr mice. It is possible that the difference of these results from 
those of other workers who have found ^butter to contain the 
essential material can be attributed to the more complete removal 
in experiments here given of lipoid-like substances from the rest 
of the food. Possibly, however, white rats do not require the 
same substances in their food that white mice require. 

It is to be noted in Experiment 14 that the fast failing animal 
using lard as the only fat in its food very quickly resumed its 
growth on the addition of an alcohol and ether extract (d) of 
egg yolk. 

To find out what portion of the alcohol-ether extract was re- 
sponsible for this growth, the various separations previously de- 
scribed were carried out and the products fed. 



Experiment 29. 



FOOD 




DATS 


WEIGHT 




Male 
gm. 

10. 
13.2 
13.4 
15.2 
13.5 
13. 
12. 
Di 


Female 


Alcohol and ether extracted casein (1) 

Extracted protein-free milk (v) 


gm. 

18. 
29.5 
23.5 
25. 


1 
15 

19 
26 
33 
40 
47 
'60 


gm. 
11. 

14.8 


Starch 


15. 


Lard 


17.2 


Xo lipoids 


16.9 


\ 


15.1 
14.2 
ed. 



l68 



Lipins in Nutrition 



Experiment 23. 








1 




WEIGHT 


FOOD 


DATS 






A 




Male 


Female 


gtn. 






gm. 


Ether and alcohol extracted casein (1) . . . 


18. 


1 


11. 


15.1 


Extracted protein-free milk (v) 


29.5 


8 


8. ! 


12. 


Starch 


26.5 
26. 
5. 


15 
22 
29 


8. I 

8.1 

6.7 


14.5 


Lard 


16.4 


Lecithin (b) 


15.5 






39 


5.5 


14.3 






43 


Died. 


14. 






50 




14. 






57 




13.6 






64 




10.5 










Died. 



Experiment 28. 



FOOD 



Ether and alcohol extracted casein (1). . 

Salt mixture (iii) 

Starch 

Lard 

Lactose 

Cephalin (a) 



DATS 



WEIGHT 



gm. 




18. 


1 


3.7 


16 


28.3 


20 


25. 


27 


20. 


34 


3. 


43 




50 

1 



Male 


Female 


gfKm 


gm. 


15. 


15.6 


16. 


17. 


15.3 


16. 


16.7 


14.9 


16. 


15. 


15.5 


16. 


13.4 


16. 



Experiment 39. 



FOOD 



Ether and alcohol extracted casein (1) 

Extracted protein-free milk (v) 

Starch 

Lard 

Third fraction (c) 





DATS 


gm. 




18. 


1 


29.5 


12 


26.5 


19 


26. 


26 


5. 


42 



WEIGHT 

gm. 

13.2 
14. 
14.3 
15. 
Died. 



C, G. MacArthur and C. L. Luckett 



169 



Experiment 26. 



FOOD 




i 
1 

DATB 


WEIGHT 




Male 
gm. 

15.6 
17.8 
19.6 
20.9 
20.4 
20.5 

23.7 
23.2 

21.4 
21. 


Female 


Ether and alcohol extracted casein (1) . . . 
Salt T^ixture (iii) 


gm. 

18. 
3.7 
28.3 
25. 
20. 
10. 


1 ■ 

5 
12 
19 
29 
33 

40 
47 
54 
61 


gm. 

17. 
17.3 


Starch 


17 9 


Lard : 


18. 


Lactose 


24 


Dried esuL yolk 


24.1 


M^ m Bw* VQQ J vasK 

• 


Young. 
20.5 
19.8 
21.2 
20.9 


Experiment 27. 








FOOD 




DATS 


WEIGHT 




Male 


Female 


Ether and alcohol extracted casein (1) . . . 
Salt mixture (iii) 


gm. 
18. 

3.7 
28.3 
I25. 
! 20. 

3. 


1 
16 

20 

27 
34 
41 
48 
68 


gm. 

17.4 
20.6 
20.5 
19.5 
18.5 
19.2 
20.8 
22.0 


gm. 

15. 
17. 


Starch 


17.4 


Lard 

Lactose 

Hot alcohol extract of egg (e and f ) 


18.8 
16.9 
18.1 
19.4 
16.1 

1 


Experiment 36. 








FOOD 


I 


DATS 


WE] 


[GHT 






i Male 


1 Female 


Ether and alcohol extracted casein (1) . . . 
Extracted nrotein-free milk (v) 


gm, 

18. 
29.5 


1 
I 

1 

: 8 
12 

19 
26 
37 
40 
60 


gm. 

14.3 
(10.5) 
14.9 
18.5 
19.7 
; 20.5 
21.5 
20.2 


gm. 

9.4 
14. 


Starch 

Tiard 

Hot alcohol-soluble nart (f ) 


23.5 
• 25. 
5. 

1 

i 


14.5 

15.4 

1 




16.6 



170 



Lipins in Nutrition 



Experiment 37. 



POOD 



DATS 



WEIGHT 



Male 



Female 



gm. 



Ether and alcohol extracted casein (1). . . ; 18. 

Extracted protein-free milk (v) 29.5 

Starch I 23.5 

I 

Lard ' 25 . 



Cold alcohol-soluble part (e) 



5. 



1 

8 '■ 
12 I 
19 , 
26 I 

33 
40 

60 i 



gm, 

15. 

16. 

15.9 

18. 

18. 

17.4 
18.2 
18.4 



gm, 

17. 
17.2 
17.6 
21. 
24. 
Young. 
18. 
20.9 
21.1 



Experiment 38. 



FOOD 







DATA 


WEIGHT 




Male ' Female 



Ether and alcohol extracted casein (1) . . . 18. 
Protein-free milk heated with alcohol 1 

day but no extract removed 29 .5 

Starch 

Lard 



Grn2 25r 



gm. 




gm, I 


18. 


1 


13.2 




12 


17. 


29.5 


17 


16. 


23.6 


24 


17.6 


25. 


45 


12. 



gm, 

12.9 

15. 

16.1 

17.7 

13.1 




DAYSO 



C. G. MacArthur and C. L. Luckett 



171 



Experiment 29 clearly shows that some substance is lacking in 
sjTithetic foods that have been extracted with organic solvents. 
Lecithin (Experiment 23) does not supply the lack nor does 
cephalin (Experiment 28), though it seems better able to main- 
tain mice than lecithin. In other experiments with cephalin none 
of the mice grew, most of them slowly declined, and some died 
after about sixty days' feeding. The third fraction (Experiment 
39), made up largely of cholesterol and fats, contained no material 
necessary for the animals' maintenance or growth. 

It is very evident that the warm alcoholic extract (Experiment 
27) causes the mouse to grow about as well as the whole dried 
egg yolk (Experiment 26). The portion of this warm alcoholic 
extract that precipitates on cooling (Experiment 36) as well as 
the portion remaining dissolved (Experiment 37) seems to supply 
the food with the missing substance. This suggests an incom- 
plete separation of the desired substance from the hot alcohol 
on cooling. That this substance is probably unstable and de- 
cidedly affected by heat is indicated in Experiment 38. In all 
cases except this last (Experiment 38) several series were run 
and the results agreed with the ones given. 

In order to check the previous results as well as to get a clearer 
idea as to the solubility of the product sought, the following 
experiments were carried out. 



Experiment 41. 



FOOD 



Pure casein (2) 

Extracted protein-free milk (v) 

Starch | 23 .5 

Lard 25 

Ether-soluble portion (h) 




WEIGHT 



Male 



Female 



gm. 


gm. 


17. 


11.7 


19. 


13.6 


20. 


13.7 


21. 


14. 


20. 


12.2 


19.5 


13. 


17.7 


11.6 


Unhcalth 


y looking 



172 



Lipins in Nutrition 



Experiment 42. 



FOOD 


gm, 

18. 
29.5 
23.5 
25. 
5. 


DATil 

1 

8 
15 
22 
29 
36 
41 


# 

WEIGHT 

Male 1 Female 


Pure casein (2) 


gm, 1 gm, 
13.2 ! 12.2 


Extracted protein-free milk (v) 


15.6 . 14.3 


Starch 


15.7 . 14 4 


Lard 


16.3 
16.4 
19.1 
19. 
Looking 


15.7 


Cold alcohol extract (a) 


(13.3) 




18.2 

18.1 
excellent. 


Experiment 44. 






FOOD 




DATB 


WEIGHT 




Male 


Female 


Pure casein (2) 

Extracted protein-free milk (v) 


gm. 
18. 

29.5 
23.5 
25. 
5. 


1 

8 
15 
22 
29 
36 
41 


gm. 

13.3 
13. 
14.5 
12.1 
9.9 
10.5 
10.7 
In poor c 


gm. 

13.5 
13.1 


Lard 


14. 
12. 
9.6 




9.7 
10.2 
jondition. 



C. G. MacArthur and C. L. Luckett 



173 



6MS20r 



DAYS 




There is little doubt that the ether solution (Experiment 41) 
containing the lecithin, cephalin, and small amounts of other 
phosphatids, cholesterol, and ordinary fats, is not essential. That 
the cold alcohol extract, containing lecithin and small amounts 
of other phosphatids, fats, cholesterol, cerebrosides, and some 
unknown substance or group of substances, is a necessary part 
of the egg both for maintenance and growth is indicated in Experi- 
ment 42. Previous experiments show that neither the lecithin, 
cephalin, cerebrosides, cholesterol, nor ordinary fats are the desired 
compound. It seems that this necessary compound is an unknown 
one, which is insoluble in ether, and either completely soluble in 
cold alcohol or destroyed by short heating with alcohol; because 
the hot alcohol extract (Experiment 44) (containing cerebrosides 
and small amounts of phosphatids and other substances), follow- 
ing the cold alcohol extract, did not keep mice in normal con- 
dition. Many of the properties of this substance suggest vitamine. 

Further work is being done to characterize this unknown sub- 
stance and, if possible, to identify it. 



174 Lipins in Nutrition 

SUMMARY. 

(^The data presented strongly indicate thatp lecithin, cephalin, 
ce?ebrosides, cholest<?rol, and fats, are dispensable parts of a food 
for mice, but 4bat a substance is present in egg yolk, insoluble 
in ether, soluble in cold alcohol, and probably easily destroyed 
by heat, that needs to be added to a synthetic food containing 
casein, starch, lactose, lard, and the salts of milk, to make it a 
complete food. 



ctric Organic Combustion Furnace 

Replaceable Unit System (Patented) 




Has met with bitCant approral by leading chemUta 

le furnaces can be used togeth^ or separately; each is 
9ped with rheostat control, allowing rapid or slow heat- 
er cooling. 

>per half of furnaces can be raised while in use, exposing 
iombuBtion tube for obserration puipoees. 

Price, complete am Illustrated, $100.00 net 

Write todiv for iUuatrated folder to 

EIMER s;^5r. AMEND 

Iquarters foi Educational and Industrial Laboratory SuppUes 



EW YOKK CITY 




PITTSBURGH, PA. 



CONTENTS 

HobkktM. Ciiapix and Wii.mer C Powick: An improved method for the 
estimation of inorganic phosphoric acid in certain tiHsues and food 
products 97 

Frgdkuic FKN(fEii: On the size and composition of the t>h>nnus gland 115 

Andrew IIitnter and Sutherland Simpson: The influence of a diet of 
marine algae upon the iodine content of sheep's thyroid 119 

Jacob Rosenbloom: A note on the distribution of mercury in the body in a 
case of acute bichloride of mercury poisoning 123 

(r. W. Raiziss and 11. Dubin: On the estimation of benzoic acid in urine.. . 125 

R. T. Woodyatt: Studies on the theory of diabetes. IV. The parallelism 
between tlio effects of the pancreas and those of metallic hydroxides on 
^ sugars 129 

Lucius L. Van Slyke and Alfred W. Hosworth: Condition of casein and 
salts in milk 135 

Caul O. Johns and Byron M. Hendrix : Researches on purines. XVI. On 
the isomeric monomethyl derivatives of 2-met.hylmercapto-4-amino-6- 
oxypyrimidine. On l-methyl-2-methylmercapto-6, 8-dioxypurine 153 

C. O. MacArthur and C. L. Luckett: Lii)ins in nutrition 161 



Tbb Journal of Bioloqical Chbiixstrt is deBigned for the prompt publica- 
tion of original investigations of a chemical nature in the biological sciences. 
It is issued monthly. Each volume consists of as many numbers as are required 
to make a total of between five and six hundred pages. 

The price of The Journal to subscribers in the United States and Canada 
is $3.00 per volume; in other countries, 13.25. Remittances should be made by 
draft or check on New York or by postal money order, payable to Tbb Journal 
OF Bioloqical Chbiiistrt, and should be sent to 2419-21 Groenmount Avenue, 
Baltimore, Md., or to The Rockefeller Institute for Medical Research, 66Ui 
Street and Avenue A, New York City. 

Dr. Dakin is at present abroad. Manuscripts may be sent to any other of 
the editors. Contributors will receive 100 copies of their papers free of charge, 
and additional copies will be furnished at cost. It is understood that, unless 
expressly agreed to the contrary, all manuscripts submitted to Tbb Journal or 
Hioi.ooical Cbbuibtbt shall not be offered for prior or simultaneous publica- 
tion elsewhere. 

The price of this number is $1.00. 



THE WAVERLY PRESS 
halt:morc. u 8. a 



VOL. XX MAIlfll, 191.-. Xo. 3 



THE JOURNAL 



1 * 



OF 

.•••■\ 

BIOLOGICAL CHEMISTRY 

rOUNDBD BT CBBIRIAM A. BBBTEB AND 8U9TAINED IN PABT BT THB CHBItTIAN A. HBBTBB 

MBlfOBlAI* rUND 

EDITED BY 

H. D. DAKIN, New York City. LAFAYETTE B. MENDEL, New Haven, Conn. 
E. K. DUNHAM, New York City. A. N. RICHARDS, Philadelphia, Pa. 

DONALD D. VAN SLYKE, New York City. 

WITH THE COLLABORATION OF 

J. J. ABBL, BftltiiiMrB, Md. J. B. LEATHES, Toronto, Canada. 

R. H. CHITTBNDBN, New Hafen, Conn. P. A. LEVENB, New York. 

OTTO FOUN, Boston, IUm. JACQUES LOEB. New York. 

WILLIAM J. GIBS. New York. A. S. LOEVENHART, Maditoa. Wit. 

L. J. HENDERSON. Cambrldfe. Hau. GRAHAM LUSK. New York. 

RBID HUNT. BoiCon, Haas. A. B. MACALLUM. Toronto. Canada. 

W. A. JACOBS, New Yoik. J. J R- MACLEOD. Clef eland, Ohio. 

WALTER JONES. BalUmoro, Md. JOHN A MANDEL, New York. 

J. H. KASTLB, Leiiaftoa. Kj. A. P. MATHEWS. Chicago. lU. 

F. G. NOVY. Ann Arbor, Mich. 

THOMAS B. OSBORNE, New Haten, Conn. 

T. BRAILSFORD ROBERTSON. Berkelej. Cal. 

P. A. SHAFFER. St. Loois. Mo. 

A. E. TAYLOR. Philadelphia. Pa. 

F. P. UNDERHILL, New Hafoa, Conn. 

V. C. VAUGHAN. Ann Arbor. Mich. 

ALFRED J. WAKEMAN, New Haf^n, Conn. 



PUBUSHED MONTHLY 
BY THB ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH 

FOR THE JOURNAL OF BIOLOGICAL CHEMISTRY, INC. 

ail9-il GREENMOUNT AVENUE. BALTIMORE, MD. 

Entered aa Beeond<e1aaa matter, August 1. 1911. at the Post OfTice at Daltlraoro. Md., under the 

Aet of Maroh 3, 1879. 



COPTBIGHT 1915 
BY 

The Journal of Biological Chbiiibtrt 



NOTE ON THE USE OF COLLOffiAL IRON IN THE 
DETERMINATION OF LACTOSE IN MILK. 

By REUBEN L. HILL. 

(From the Department of Physiology and Biochemistry^ Cornell University 

Medical College, Ithaca,) 

(Received for publication, January 1, 1915.) 

The method described below has the advantage of being within 
the reach of the ordinary laboratory student; it requires compara- 
tively little time and gives very accurate results. ^ 

In clarifying the milk, a 10 per cent solution of colloidal iron 
(dialyzed ferric hydroxide) is used. By adding the proper amount 
of colloidal iron, all the proteins of the milk are completely pre- 
cipitated and can be rapidly filtered off leaving a perfectly clear 
colorless filtrate. 

The method is as follows: To a 10 gram sample of milk, which 
has been diluted to about 25 cc, about 3 cc. of a 10 per cent 
solution of colloidal iron are added. The amount of colloidal iron 
necessary depends upon the composition of the milk and can be 
accurately determined by adding the last portion drop by drop, 
and agitating after each addition. 

If the precipitation is complete, a clear supernatant liquid 
separates out from the flocculent precipitate; if too little has been 
added, the supernatant liquid will appear milky; if too much, it 
will have a reddish tinge. 

The sample is next filtered into a 100 cc. volumetric flask, and 
the precipitate thoroughly washed with distilled water until the 
filtrate and washings aggregate about 100 cc. The flask is then 
filled to the mark and the percentage of lactose determined by 
Benedict's quantitative method.^ About 16 cc. of the diluted 
sample will be required to reduce completely 25 cc. of Benedict's 
quantitative solution. 

1 S. R. Benedict: Jour. Am. Med. Assn., Ivii, p. 1193, 1911. P. B. Hawk: 
Practical Physiological Chemistry, 4th edition, Philadelphia, 1912, p. 386. 

175 

THE JOCTBNAL OF BIOLOOICAL CHSUtSTRY. VOL. XX, NO. 3 



176 Determination of Lactose in Milk 

A very convenient method of analysis is given by Cole,- in 
which a 4 ounce flask is used instead of an evaporating dish. The 
wide mouthed Jena 150 cc. flat bottomed flasks are very conven-. 
ient for the determination. The flask is fitted into the 2.5 inch 
ring of a retort stand, and the height above the Bunsen byrner so 
arranged that the contents of the flask will be kept briskly boiling 
with a small flame. Two flasks can be run simultaneously from 
the same stand. 

Three to four grams of anhydrous sodium carbonate are dis- 
solved, by means of heat, in 25 cc. of twice diluted' Benedict's 
solution, to which a little powdered pumice has been added. 
About 14 cc. of the sugar solution are then rapidly added from a 
burette. Boiling is continued for at least one-half minute before 
the addition of more lactose solution. 

When reduction is complete the supernatant liquid will have a 
slight yellowish tinge to which the blue color very slowly returns. 
If the end-point has been underestimated, it will have a blue or 
greenish tinge that rapidly becomes bluer. With a liltle prac- 
tice, and by adding the last portion a drop at a time, and boiling 
one-half minute after each addition, the end-point can be deter- 
mined to within one drop. 

Twenty-five cc. of Benedict's quantitative solution are com- 
pletely reduced by 0.0676 of a gram of anhydrous lactose. Since 
the milk has been ten-fold diluted, 0.0676 divided by the number 
of cc. of diluted lactose solution used, multiplied by ten, will give 
the percentage of lactose in the milk. If 16.1 cc. of lactose solu- 
tion were required, then 16.1: 0.0676 :: X: 10 = 4.20 per cent. 

That veiy accurate results can be obtained by using this method 
may be seen from the following tables. Table I shows the com- 
parison between duplicate samples of different milks. Table II 
shows the effect of the addition of the small quantities of lactose* 
to the milk before analysis. 

- S. W. Cole: Practical Physiological Chemisiryj St. Louis. 101 *, p. .i3. 

3 1 prefer to (Ulute 25 cc. of Benedict's solution to at least 50 cc, for I 
obtain a more accurate end-point with dilute than with concentrated 
solutions. 



Reuben L. Hill 

TABLE I. 
Comparison betweert duplicate sampler of different milks. 



177 



S« AMPLE NO. 



ANHYDROUS LACTOSE 
IN MILK 1 



1 
2 
3 



per cent 

4.08 
4. II 
4.10 



LACTOSE IN MILK 2 



per cent 



4.19 
4.20 
4.20 



LACTOSE IN MILK 3 

per cent 

4.23 
4.20 
4.22 



TABLE II. 
Effect of the addition of lactose to 10 cc. samples of milk. 





SAMPLE, TEN TIMES 








A\MPLE NO. 


DILUTED, REQUIBED 

TO BXDUCE 25 CC. 

or BENEDICT'S 

QUANTITATIVE 

SOLUTION 


ANHYDROUS 

LACTOSE IN 

SAMPLE 


ANHYDROUS 
L.ACTOSE ADDED 

gm. 


ADDED LACTOSE 
RECOVERED 




CC. 


gm. 


! per cent 


Milk 4 




\ 






1 


16.7 


0.405 


None 


1 


2 


14.9 


0.454 


0.050 


98 


3 


15.0 


0.451 


0.050 


92 


Milk 5 










1 


16.1 


0.420 


None 




2 


14.4 


0.469 


0.050 


98 


Milk 6 










1 


16.4 


0.413 


None 






7.5 


0.902 


0.505 


07 


3 


10.2 


0.664 


0.2525 


99 


4 


10.1 


0.670 

1 


0.2525 


100.8 



SPECTROSCOPIC INVESTIGATION OF THE REDUCTION 
OF HEMOGLOBIN BY TISSUE REDUCTASE.^ 

By DAVID FRASER HARRIS, M.D., D.Sc, F.R.S.E., and HENRY 
JERMAIN MAUDE CREIGHTOX, M.A., M.Sc., Dr.Sc. 

(From the Laboratories of Physiology of Dalhousie Universityy Halifax^ 
N. S., and the Laboratories of Chemistry of Sivarthviore 

College f Swarthmore.) 

(Received for publication, January 5, 1915.) 
I. INTRODUCTION. 

In our communication, *' Studies on the reductase of liver and 
kidney. Part I," we showed* that ox liver juice two days old was 
able to reduce methemoglobin to the well known two-banded 
oxyhemoglobin, within ten minutes at 40°. This methemoglobin 
was prepared from *' hemoglobin scales, soluble. Merck." This 
observation we have confirmed and extended. It has been found 
that fresh cat liver juice will reduce methemoglobin not only to 
the state of oxyhemoglobin, but to that of the completely reduced 
pigment (Hb). The juice used was press juice made by the tech- 
nique employed in our previous work, and the methemoglobin 
was procured by exposing cat defibrinated blood (diluted 1 in 25) 
to the air in an ice box for several days. A mixture of 10 cc. of 
juice with 25 cc. of the nlethemoglobin placed in the thermostat 
at 40** showed within three minutes the absence of the band in 
the red region and the appearance of the two well marked bands of 
oxyhemoglobin; while in six minutes more these two bands had 
gone, to be replaced by the single band of reduced hemoglobin. 
The changes visible to the naked eye corresponded with the 
spectroscopic appearances; the mixture originally of a rusty 

^ Part of the expenses of this research were met from the residue of a 
grant from the Government Grant Committee of the Royal Society previous- 
ly acknowledged. 

* D. Fraser Harris and H. J. M. Creighton: Proc. Roy. Soc, Series B, 
Ixxxv, p. 487, 1912. 

179 



i8o Reduction of Hemoglobin by Tissue Reductase 

brown passed through a pink to become the purple colour so 
characteristic of solutions of fully reduced blood. We found that 
this fully reduced pigment could be oxidised to two-banded oxy- 
hemoglobin on shaking with air, and that, when placed once more 
in the thermostat, it became again reduced to the one-banded 
condition. 

The work about to be described we undertook because we 
thought it well to investigate the behavior of reductase in the 
presence of oxyhemoglobin, which is the natural source of the 
oxygen with which the enzyme deals in the living tissues. Such 
chemical substances as we had previously used to demonstrate 
reduction by tissue juices must be all, more or less, poisonous to 
11 tissue enzyme, so that by using oxyhemoglobin we have the 
reductase in the presence of what might be called a natural sub- 
strate. Further, by using oxyhemoglobin to be reduced by tissue 
juices, we thought that certain data might be obtained which 
would shed hght directly on the general problem of tissue respira- 
tion, a process which is being more and more thought of as enzymic 
in its essential features. We saw no reason to doubt that fresh 
tissue juices could at least bring about such reductions as do the 
intact tissues themselves. Oxyhemoglobin is reduced, though 
not completely, in one capillary transit of the circulating blood, 
so we thought it well to try to ascertain the time relations of the 
complete reduction of oxyhemoglobin to hemoglobin, the temper- 
ature coefficient of the change which we ascribe to a tissue enzyme, 
the rate of decay in activity of this enzyme, and any other data 
likely to be of service in a conception of the process of internal 
respiration. 

II. TECHNIQUE. 

The late Professor Yeo in 1885 had employed* a spectroscopic 
method to demonstrate the reduction of manunalian oxyhemo- 
globin by the activity of the beating frog heart. Certain observers* 
had immersed pieces of tissue in solutions of oxyhemoglobin, 
and after varying intervals of time had examined the solutions 
spectroscopically. 

» G. F. Yeo: Jour. Physiol., vi, p. 93, 1885. 

* J. G. Mackcndrick: Proc. Roy. Soc. Edinburgh^ xxii, p. 201, 1807-98. 



D. F. Harris and H. J. M. Creighton i8i 

In the replacing of the two bands of oxyhemoglobin by the one 
liand of fully reduced hemoglobin, we have an optical end-point 
which is not affected by any personal factor. But the disappear- 
ance of the two bands is from a very early period accompanied by 
the appearance of a fainter band between them, i.e., in the intra- 
linear green, so that at a certain stage of reduction we have the 
single band of reduced hemoglobin flanked by the remains of the 
two bands of the oxyhemoglobin. We have not considered the 
pigment fully reduced until these margins of the single band have 
completely disappeared, so that in this way we contrived to have 
an end-point of considerable definiteness. We used the small 
direct vision spectroscope. It was early evident that owing to 
the opacity of the mixture of tissue juice and diluted blood in 
even a narrow test-tube, the spectrum of the oxyhemoglobin 
could not be seen sufficiently distinctly to be certain of the end- 
point. We found that small vessels made by drawing out short 
lengths of glass tubing (6 mm. diameter) to a conical end answered 
our purpose very well; for if in doubt as to the persistence of the 
two bands, when we directed the spectroscope across the tube 
itself, it was only necessary to look through the conical part of 
the tube where the mixture was sufficiently transparent to observe 
a distinct spectrum. Another objection to the use of even a 
narrow test-tube as an observation tube was, that if we filtered off 
some of the mixture of juice and diluted blood in order to obtain 
a clear liquid,* any reduced hemoglobin was certain to be rapidly 
reoxidised to the two-banded condition. This source of error was 
felt to be so serious that we early abandoned the method of filter- 
ing, and decided that the spectrum must be examined in a mixture 
of juice and diluted blood to which oxygen could not gain access. 

All animals used were decapitated during chloroform anesthesia. 

The dilution of blood found most suitable for spectroscopic 
[)urposes in a semi-opaque mixture was 10 cc. of defibrinated 
blood diluted with water to 250 cc. This dilution, viewed through 
the conical observing tube, showed distinctly the two bands of 
oxyhemoglobin separated by some intralinear green Ught in such 
a way that the subsequent filling up of this coloured region by the 
one band of the reduced hemoglobin was distinctly noticeable. 
Complete reduction was considered to have taken place at the 
first moment when the two bands had vanished, and when they 



i82 Reduction of Hemoglobin by Tissue Reductase 

were no longer distinguishable as ribbed margins to the single 
band. 

When only the Uver was required for juice, it was perfused 
through the portal vein with tap water at 40° until bloodless. 
When muscles, heart, pancreas, other viscera, or cortex cerebri 
were needed, the perfusion cannula was tied into the origin of the 
aorta. The juice from the tissues, whether crushed in the juice 
press or macerated with sand in a mortar, was received and kept 
under toluene. A mixture of 10 cc, of juice with 25 cc. of diluted 
blood (1 in 25) was employed in all cases unless otherwise stated. 
This mixture in a conical tube was always covered with a layer of 
toluene to a depth of 1 cm. 

m. EXPERIMENTAL. 

A, Redvdion with liver juice. 

(a) Cat liver juice and oxyhemoglobin, A mixture of freshly 
prepared cat Uver juice and its diluted blood was placed in the 
thermostat at 40° at 4.15 p.m., and was examined with the spec- 
troscope every minute. At 4.28 it was found to be one-band?d. 
The boiled control tube remained two-banded at the end of this 
day's scries of observations, three hours later. On vigorously 
shaking the reduced mixture with air, it at once became dis- 
tinctly two-banded. A second mixture of liver juice and blood in 
the usualproportions was placed in the thermostat at 40** at 4.33 p.m. 
At 4.44, i.e., in eleven minutes, it was found to be one-banded. It 
was capable of reoxidation, and on being returned to the bath was 
found to undergo reduction in the same time as the trst, eleven 
minutes. Other observations on the same day with juice not 
more than four hours old gave similar results. 

In order to eliminate every source of error, we thought it well 
to determine whether dextrose could possibly be the agent active 
in reducing the oxyhemoglobin, as dextrose is a constant constit- 
uent of the liver of all but fuUy starved animals. Accordingly 
we mixed 2 cc. of a 2 per cent solution of dextrose with 4 cc. of 
the diluted blood used in these experiments, and kept the tube at 
40° for 105 minutes, at the end of which time the solution was still 
two-banded. Since fresh liver juice can reduce oxyhemoglobin 
at 40° certainly within fifteen minutes, the reduction must b(» 
due to something in the juice other than dextrose. 



D. F. Harris and H. J. M. Creighton 183 

Ebcactly similar observations with bile convinced us that none 
of the reduction of oxyhemoglobin by liver juice was due to any 
residual bile that may have been in the liver. 

(6) Rabbit liver juice. Observations with mixtures of rabbit 
liver juice and rabbit blood pelded results similar to those just 
tlescribed. 

(c) Pigeon liver juice. Our experiments with this juice showed 
us that it contained the most active reductase we have yet met 
with. A mixture of fresh pigeon liver juice and pigeon diluted 
blood in the usual proportions showed complete reduction so 
rapidly that there was no time to place the mixture in the ther- 
mostat at 40°; that is, at room temperature (17°) the reduction 
was effected at once. A boiled control showed no change at the 
end of several hours. 

This great energy of reduction as displayed towards oxyhemo- 
globin we found to be equally well exhibited towards soluble 
Prussian blue. 1 cc. of the liver juice caused 2 cc. of a 0.5 per cent 
solution of soluble Prussian blue to fade instantly at room tem- 
perature. More of the Prussian blue solution was added until 
8 cc. had been poured in; the blue colour faded to the natural fawn 
liver colour after each addition of the pigment. The boiled con- 
trol showed no change in colour after similar additions of soluble 
Prussian blue solution. The entirely bleached mixture was 
restored to a blue colour on the addition of hydrogen peroxide. 

(d) Frog liver juice. The juice of frog liver was tested 
against frog blood both at room temperature (20°) and at 40°. 
At the former, the oxyhemoglobin was completely reduced 
within thirty-seven minutes; at 40° the reduction was complete 
in almost the same time (thirty-five minutes). A large amount 
of very dark pigment in the liver rendered the spectroscopic 
observations difficult. 

(e) Fish liver juice. The non-perfused liver of a mackerel, 
caught about twenty-four hours previously, was crushed in the 
juice press: a homogeneous, viscous, Uvid juice was obtained. 
This viscous juice was mixed with two-thirds of its volume of 
water in order to have a juice of suitable fluidity. The blood of 
the liver was seen to be already reduced in this juice. Of this 
fluid Uver juice, 2 cc. were mixed with 4 cc. of the fish blood 
(dilution 1 in 25) ; reduction took place at once at air temperature 
(20°). 



184 Reduction of Hemoglobin by Tissue Reductase 

In a second experiment 2 cc. of the liver juice were mixed with 
8 cc. of the blood solution; this mixture was introduced into the 
observation tubes, one of which was placed in the bath at 40° 
and the other allowed to remain at room temperature. The 
oxyhemoglobin in the tube at 40° was found to have become one- 
banded at the end of one minute, while that in the tube at air 
temperature required two minutes for complete reduction. 

In another experiment 1 cc. of the Uver juice was mixed with 
8 cc. of the blood solution. Observ^ation tubes containing por- 
tions of this mixture were placed in the batji at 40° and in a bath 
at room temperature. At the end of two and a half minutes 
reduction had taken place at 40°, while seven and a half minutes 
were required to effect this at 20°. Observations made with fish 
blood solution and liver juice which had been boiled for a minute 
showed no evidence of reduction at the end of several hours. 

The reducing power of the reductase in this Uver juice was 
corroborated by our finding that 1 cc. of it completely decolourized 
at 20° 10 cc. of a 0.05 per cent solution of soluble Prussian blue. 

(/) Ex'perimenis on crossed reductions. By crossed reduction 
we mean the action of reductase from any animal A on the oxy- 
hemoglobin of any other animal B. It was highly probable that 
this would occur, seeing that frog heart, for instance, can reduce 
rabbit or ox dilute defibrinated blood in physiological perfusion 
(experiments. The late Professor Yeo had shown^ by the use of 
the spectroscope that the frog heart could reduce mammahan 
oxyhemoglobin to the one-banded condition. 

With cat Uver juice at 40°, it was found that this substance 
reduced pigeon Wood in half the time (three and five-tenths min- 
utes) in which it reduced cat blood (six and five-tenths minutes). 
Cat Uver juice reduced frog blood in six minutes at 40°, or in the 
same time in which it reduced cat blood. Similarly, pigeon liver 
juice reduced cat blood in about the same time that it reduced 
pigeon bloocL In exi)eriments on pigeon muscle juice, we found 
that it reduced pigeon blood more rapidly than it did cat blood. 

^ Veo: loc. cit. 



D. F. Harris aad H. J. M. Creighton 185 

B, Reduction with niusde juice, 

(a) CcU musde juice. Contrary to our expectations it was not 
difficult to obtain a press juice from cat muscles. A highly 
colloidal juice, which coagulated on boiling, was received under 
toluene. This juice mixed in the routine proportions with di- 
luted blood and placed in the thermostat at 40° showed little 
reduction at the end of half an hour. A boiled control showed 
none whatever, but the oxyhemoglobin in this case underwent a 
change which we can only call fading; for its colour became a pale 
brown and its two bands became fainter and fainter until they 
were just discernible. Neither at room temperature nor at 40*^ 
did cat muscle juice (two hours post mortem) fully reduce cat 
blood. The fading of the oxyhemoglobin was, we think, largely 
due to adsorption, whereby nearly all of it was removed from so- 
lution. One factor in the failure to reduce the oxyhemoglobin may 
have been that the juice was prepared from muscles two hours 
after death. 

This muscle juice tested with soluble Prussian blue effected 
only a very slight reduction of that pigment, for the mixture of 
the muscle juice and Prussian blue did not become colourless or 
even gray, but only a pale green. The boiled control, however, 
remained blue, although it had adsorbed some of the colour. 

(6) Pigeon muscle juice. The non-perfused pectoral muscles 
of a pigeon, before the animal heat had left them, were disinte- 
grated, ground up with sand and a little water and filtered through 
cheese-cloth. Of this filtrate 2 cc. viewed in the conical obser- 
vation tube at room temperature showed that the residual blood 
of the musclejjras already reduced. Two cc. of the muscle juice 
were mixed with 2 cc. of pigeon blood at room temperature, and 
in two minutes the oxyhemoglobin was found completely reduced. 
On the other hand, the boiled control remained two-banded for 
many hours. As estimated by its action on soluble Prussian blue, 
the muscle juice was an active reducer, but not so active as the 
liver juice of either the cat or the pigeon. 

(c) Cardiac muscle juice. The still warm hearts of four kittens 
were crushed in the juice press and ground up with sand and a 
little water. The mixture was filtered through cheese-cloth. 
The usual mixture of juice and diluted blood did not at any time 



1 86 Reduction of Hemoglobin by Tissue Reductase 

show the disappearance of the two bands of oxyhemoglobin; in- 
deed, there was little difference between the active tubes and the 
l)oiled controls. Quite different was our experience with the 
heart juice of a small rabbit. The entire heart while still warm 
was ground up with sand and a little water, and some of the fil- 
trate through cheese-cloth was at once mixed with some of the 
animal's defibrinated diluted blood (1 in 25). This was reduced 
in less than three minutes at room temperature, while a boiled 
control remained unchanged at the end of two hours. A frog 
heart crushed and mixed with diluted frog blood at room tem- 
perature had not reduced the oxyhemoglobin at the end of thirty 
minutes. 

C Reduction with the juices of other organs. 

(a) Kidney. The eight non-perfused kidneys of four decapitated 
kittens were crushed and the juice was prepared in the usual way. 
The blood was defibrinated and diluted, though not quite to such 
an extent as was the blood employed in the preceding experiments. 
A mixture of juice and blood made in the usual proportions 
showed complete reduction in fifty-seven minutes at 40°. On 
shaking the tube with air, the reduced pigment was reoxidised to 
two bands as distinct as they had been at first, but on returning 
the mixture to the bath its reduction was effected only with very 
great slowness. 

(6) Stanmch. The tissues of the wall of the stomach of a cat 
were disintegrated while still warm and ground up with sand and 
a little glycerine. Some of the resulting juice was mixed with 
diluted cat blood and placed in the thermostat at 40*^. The juice 
slowly separated from the blood solution, making it impossible 
to maintain a homogeneous mixture; but where the juice and 
blood came in contact, reduction had clearly taken place in less 
than half an hour. 

(c) Pancreas. Only one observation was made with disint-e- 
grated cat pancreas and blood, and it was negative. 

(d) Cortex cerebri. Some of the fresh cortex cerebri was ground 
up with a little glycerine. A mixture of 1.5 cc. of the juice and 10 
cc. of diluted cat blood was placed in the thermostat at 40°. In 
forty-seven minutes the oxyhemoglobin was fully reduced, whereas 



D, F. Harris and H. J. M. Creighton 187 

that in the boiled control was still two-banded at the end of several 
hours. 

The reducing power of the cerebral tissue was confirmed in an 
experiment with soluble Prussian blue. A mixture of 1.5 cc. of 
the juice with 10 cc. of 0.1 per cent solution of Prussian blue was 
reduced within a minute to the gray condition, more blue was 
added, when the mixture again became gray; a third time blue 
was added with the same result. A mixture containing 1.5 cc. 
of boiled juice and 10 cc. of the Prussian blue solution remained 
blue-green. 

D. Effect of temperature on the reducing power of reductase, and the 

rate of decay of its activity. 

In the following experiments the reductase contained in cat 
liver juice was employed. 

The reducing power of reductase, at different temperatures, 
was measured by determining the time required for the reduction 
of two-banded hemoglobin to the one-banded condition. This 
procedure was also employed to determine the variation in the 
reducing power of reductase with time. The defibrinated cat 
blood used in these measurements was kept in an ice box, where 
it was found to have undergone no apparent deterioration at the 
end of eight days. The blood was always oxygenated by shaking 
with air, before being used for the preparation of the hemoglobin 
solutions. In the following experiments 10 cc. of the liver juice 
were mixed with 25 cc. of oxyhemoglobin solution, prepared by 
diluting 25 cc. of defibrinated cat blood to 250 cc. with water. 
This mixture was then divided into several parts, which were 
introduced into the observing tubes, and these were then quickly 
immersed in different thermostats. The temperatures of the 
thermostats were kept constant to ± 0.2^. 

Table I gives the time required at different temperatures for 
the reduction of a solution of oxyhemoglobin, prepared from fresh 
defibrinated cat blood, by fresh cat liver juice (half an hour post 
mortem). 

In Table I column 6 gives the temperature coefficient of the 
process of reduction for intervals of 10°. These values have Ix^en 
obtained by dividing the average time required for complete 



1 88 Reduction of Hemoglobin by Tissue Reductase 



reduction at any one temperature by that required at a tempera- 
ture 10° higher. Column 7 gives the reducing power of the 
reductase of cat liver juice in units obtained by multiplying- l^y 
100 the reciprocal of the time required for reduction at a particular 
temperature. The change in the reducing power of the reductase* 
of cat liver juice with temperature is shown graphically in Figure 1 . 
In addition to the foregoing, measurements of the reduction of 
oxyhemoglobin by reductase were carried out at 60° and 70°. At 
the former temperature one and three-quarters minutes were re- 
quired to effect reduction, and at the latter, one and one-quarter 
minutes. These measurements above 60° are probably of little 









TABLE I. 


■ 






31 


TIME RSQUIRED FOR COMPI^BTE REDUCTION * 








D 




OF HbOt 




TEMPERATURE 


REDUCING POWKR 


% 








COEFFICIENT 


OF 


RBDUCTA8I: 


PS 

u 

M 


Experiment I Experiment 

II 

a b i 


Average 
time 


100 


•r. 


min . 


min. min. 


min. 











98 


98 


98 


2.05 
1.91 
2.07 
2.00 
1.86 




1.02 


10 


36 


38 


37 




2.70 


20 
30 


22 
10 


21 • 15 

10 8 


19.3 
9.3 




5.17 
10.72 


40 


5 


5 4 


4.7 




23.00 


oO 


2.5 


2.5 2.5 


2.5 




40.00 


55 


1.75 


1.75 


1.75 






57.14 



value owing to the decomposition of the oxyhemoglobin; for, as 
one of us has shown,® at temperatures above 60^ the two bands 
of oxj^hemoglobin rapidly become hazyJ 

Measurements were made from day to day of the reducing 
power of the reductase of the liver jaice on an oxyhemoglo})in 
solution made from defibrinated cat blood of the same age as the 
juice. These determinations were carried out at different tem- 
peratures, in order to ascertain whether the ratio of the times 

^ D. Frascr Harris: Froc, Roy. Soc. Edinburgh, xxii, p. 192, 1897-98. 

^ The temperature coefficient from 0° to 55** (Table I) varies at both 
extremes of this range in a manner similar to that observed by Van Slyke 
and Cullen in the case of urease (this Journal, xix, p. 174, 1914). The simi- 
larity in the behavior of the two enz\Tnes indicates a probable likeness in 
their mode of action. 



D. F. Harris and H. J. M. Creighton 189 

required for reduction at any two temperatures varied with the 
age of the reductase. The results of the observations, which are 
given in Table II, show a marked decrease in the activity of the 
enzyme from day to day. The times given in this table for reduc- 
tion represent the means of two experiments which seldom dif- 
fered by more than a few seconds and which, in most cases, were 
identical. 

Tile change with temperature in the reducing power of the 
reductase of cat liver juice of different ages is shown graphically 
in Figure 1. It will be seen from the shape of the curves that the 
ratio of the times required for reduction at any two temperatures 
undergoes little or no variation with the age of the reductase. 

TABLE IT. 



a: 

< 

H 
H 



24 HR. OLD JUICE 



44 HR. OLD JUICE 



02 HR. OLD JUICE 



18S HR. OLD JUICE 



Time 

required 

for the 

ireduetion 

of HbOs 



Reducing I xime Reducing 
PO^?'.o' ! required Powwof 
the juice I for the ^^^^jJo'^® 



100 



reduction 
I of Hb0» 



•C. 

10 
20 
30 
40 
50 
55 



nun. 

41 

19.3 


5.4 
3.3 
2.3 



'<?• 



2.44 
5.18 
11.11 
18.52 
30.3 
43.5 



mtn. 

55 
28 
12 

6.5 

4 

3 



1.90 

3.57 

8.33 

15.38 

25.0 

33.3 



Time 

required 

for the 

reduction 

of HbOa 



min. 

124 
58 
30 
15 

8.5 

5 



Reducing 

power of 

the juice 

100 



0.81 
1.72 
3.33 
6.66 
11.76 
20.0 



Time 
required 
for the i 
"reduction i 
of HbOt ; 



min. 



230 
94 
39 
27 



Reducinc 

power of 

the Juice 

WO 



0.43 
1.06 
2.56 
3.70 



Figure 2 shows the change of the reducing power of the reductase 
of cat liver juice with age, for several temperatures. In this 
figure two sets of curves are given: the reducing power of the 
reductase expressed in the units given in the tables; and the 
logarithms of these units are plotted on the axis of ordinatcs against 
the age of the enzyme on the axis of abscissae. The curves found 
by plotting the logarithms of the reducing powers against the age 
of the reductase are straight lines over the whole period of time. 

The numbers in the foregoing tables show, as is to be expected, 
that the reductase of cat liver exhibits as an essential character- 
istic a marked sensitivity towards changes in temperature; and, 
analogous with the majority of chemical reactions, the velocity 
of the reduction of oxyhemoglobin to the one-banded condition 



THEJOURXAL OF BIOLOGICAL CHEMISTRT, VOL. XX, .VO. 'i 



igo Reduction of Hemoglobin by Tissue Reductase 



60 



^ 



P^40 
liJ 

c!> 

Z 

o 

D 
U 

tr 



10 



\ 














\ 


i 




A: Fresh Juice 
B: 44 Hour Old Juice 
C: 9Z *^ '' ^' 
D;l88 " " ^ 




\ 


\ 






\ 


Aj 












\ 


^f 


\ 














^...^ 




^ 


g___ 





60* 



SO* 



40* ^* 20* 10* 

Temperature 

Fig. 1. 



©• 



-10* 




Time in Hours 
Fig. 2. 



D. F, Harris and H. ]• M. Creighton 191 

by the enzyme increases with temperature. Between 10° and 40® 
the velocity of reduction is approximately doubled for every 10° 
rise in temperature, the temperature coefficient being about 2. 
Above 40° the temperature coefficient and the consequent accel- 
eration of the velocity of reduction decrease rapidly with the rise 
ill temperature. Between 50° and 60° the temperature coeffi- 
cient has been found to be 1.43. Although, in general, the tem- 
perature coeflScients of chemical reactions decrease slightly with 
increase in temperature, the decrease in the values obtained for 
the reduction of oxyhemoglobin by reductase, at temperatures 
above 40°, is much greater than would be the case in an ordinary 
chemical reaction. It is probable, therefore, that above 40° some 
new influence makes itself felt. It has been shown by one of us** 
that the optimal temperature of reductase lies between 42° and 40°. 
In view of this, it is likely that, at temperatures between 40° and 
60°, the acceleration of the velocity of reduction due to increase 
in temperature is to a certain extent counteracted by a partial 
inliibition or destruction of the enzj^me, the result being a rela- 
tively large decrease in the value of the temperature coefficient. 
Below 10° the velocity of reduction of oxyhemoglobin diminishes 
very rapidly with decrease in temperature, and, indeed, below 
0° it is so extremely slow that the inhibition of the enzyme 
may be regarded as almost complete. That cold does not per- 
manently inhibit the reductase is shown by the fact that when 
liver juice, which has been kept at 0° for several hours, is mixed 
with a solution of oxyhemoglobin at 40°, reduction proceeds at 
the usual velocity. This substantiates similar observations pre- 
viously made* with liver juice and soluble Prussian blue. 

The decrease in the reducing power of cat liver juice at any 
temperature, with increase in the age of the juice, is douV)tless 
due to a fall in the activity of the enzyme, a phenomenon which 
is common to most enzymes in solution. Since the same prepara- 
tion of liver juice, which was kept at room temperature, was used 
in all the foregoing temperature measurements, the fall in the 
activity of the reductase should occasion the same percentage 
<l(^crease in the reducing power of the juice, independent of the 

* Harris: Biochem. Jour., v, p. 158, 1912. 

* Plarris and Creighton: Proc. Roy. Soc, Series li, Ixxxv, p. 491, 1912. 



192 Reduction of Hemoglobin by Tissue Reductase 

temperature at which the reduction was carried out. The sha^x? 
of the curves in Figure 2 confirms this. The reductions carried 
out at 30°, 40°, 50°, and 55° show that, at the end of eight days, 
the reducing power of the juice had decreased to 5-6 per cent of 
its original value. This, of course, indicates a corresponding 
decay in the activity of the reductase. 

The character of the curves in Figure 2, portraying the decrease 
in the reducing power of reductase with the age of the juice, shows 
that the decay of the activity of the enzyme follows the loga- 
rithmic law. 

The decay in the activity of reductase is of interest. Shaklee^^ 
has found that the decay in the activity of pepsin at 37° is given 
by the equation, 

X 



tia-x) 



fc, 



where x is the percentage quantity of pepsin rendered inactive, a 
the original quantity, t the time, and k a constant (= about 0.5). 
After maintaining pepsin at this temperature for twelve days, it 
was found to contain only 14 per cent of the active enzyme. This 
formula has been applied without success (compare Table III) 
to our reduction measurements. In applying this formula the 
reducing power of the fresh juice has been taken as 100 per cent, 
and it has been assumed that the reducing power of the reductase 
is proportional to the amount of active enzyme present. 

In order to ascertain whether the decay in the activity of reduc- 
tase with age follows the unimolecular law, the results given in 
the foregoing tables have been used in the equation, 

log =A;, 



0.4343 i •'a-x 

to calculate the value of the velocity constant, fc. Since the 
activity of the enzyme is proportional to its reducing power, a, 
the initial activity, may be represented by 100 per cent, and x, 
the decrease in activity at the end of time, i, may be represented 
by the percentage decrease in reducing power at the end of this 
time. The values obtained for k with the unimolecular and 
Shaklce^s equations are given in Table III. 

»o A. O. Shaklee: Ceniralbl.f. Physiol., xxiii, p. 4. 1909-10. 



D. F. Harris and H. J. M. Creighton 193 



Although the values for the unimolecular constant given in 
Table III show considerable differences, they are sufficiently con- 
stant to indicate that the decay of activity of reductase undoubt- 
edly follows the unimolecular law. When the inaccuracies of 
the method employed for measuring the reducing power of reduc- 
tase, and the uncertainties of the initial and end-points of the 









TABT.K III. 






1 

\M9 


1 
1 


Jl- ' 


1. ^ 1 « 




jaJE 


^ » 

per cent 


1 
1 

per cenl 


t{a - X) 


X 


ir«. 











100 





Temperature 55* 






24 


76.2 


23.8 


0.0130 


0.0113 




44 


58.3 


41.7 


0.0162 


0.0141 




92 


35.0 


65.0 


0.0202 


0.0097 




188 


6.5 


93.5 


0.0765 


0.0146 






Mean: 0.0132 







100 





Temperature 50** 






24 


75.7 


24.3 


0.0133 


0.0116 




44 


62.5 37.5 


0.0136 


0.0107 




92 


29.5 


70.5 


0.0259 


0.0133 




188 


6.5 


93.5 


0.0765 


0.0146 








Mean:0.C134 







100 





Temperature 40® 






24 


80.0 20.0 


0.0104 


0.0093 




44 


66.6 


33.4 


0.0114 


0.0093 




92 


29.0 


71.0 


0.0266 


0.0135 




188 


4.8 


95.2 


0.1055 


0.0162 






Mean: 0.0121 





process are considered, the divergencies in the values of the velocity 
constant are not surprising. On the other hand, the continual 
increase of k with time, when calculated by means of Shaklee's 
formula, makes it evident that this equation cannot be employed 
to determine the decay in the activity of reductase. As is to be 
expected, the same value is obtained for the unimolecular velocity 
constant at different temperatures; for x represents the decrease 



194 Reduction of Hemoglobin by Tissue Reductase 

in the reducing power of the liver juice at room temperature, the 
temperature at which the juice was kept. The variations in the 
room temiDcrature would partly account for changes in the value 
of k. Since it is evident that the decay in the activity of reduc- 
tase follows the unimolecular law, and that the activity of thc^ 
enzyme is proportional to its reducing power, the percentage of 
active reductase present in a juice of any age may be calculated 
by means of the unimolecular equation. 

IV. DISCUSSION ON THE PHYSIOLOGICAL P.EARING OF THK 

EXPERIMENTS. 

We desire, in the first place, to point out the exacting nature* of 
the chemical task set before tissue reductase in our experiments. 
In no case did we report full reduction until the two bands of 
oxyhemoglobin had given place to the one band of the reduced 
pigment. In other words, we forced the reductase <to accomplish 
in vitro a degree of reduction of the blood pigment, which it is 
never called upon to do in vivo within one circuit of the blood 
through any capillary district. The blood leaving a capillary dis- 
trict is stated to possess about two-thirds of the oxygen it had on 
entering it; that is, it is far from being completely reduced. Not 
until it has made several circuits is the oxyhemoglobin all reduced 
and the tissues said to be asphyxiated. The dissociation of the 
oxygen from the hemoglobin is the event of physiological impoi- 
tance in the depths of the tissues. 

Recently much work has been done on the problem of th(» 
physicochemical mechanism of the liberation of the oxygen from 
oxyhemoglobin. Barcroft" has investigated the influence of the 
presence of carbon dioxide in the blood as a factor in the dis- 
sociation of the oxygen from oxyhemoglobin. In cold-blooded 
animals he holds that it is a most important factor; but surely 
in all kinds of animals the oxygen avidity or reducing power 
of the tissues is the chief factor. Starling, quoting Barcroft. 
writes: '*In cold-blooded animals the dissociation of oxyhemo- 
globin with the setting free of oxygen must be largely conditioned 
by the rise of carbon dioxide tension in the tissues, since at th(» 
normal temperature of these animals the evolution of oxygen from 

" J. Harcroft and W. O. R. King: Jour. Physiol., xxxix, p. 377, 1900 10. 



D. F. Harris and H. J. M. Creighton 195 

hemoglobin is extremely slow." We submit that the part played 
by reductase in this tissue respiration is the main factor. At 
room temperature, cold-blooded animal tissue juices possess 
reducing power. No doubt carbon dioxide is being evolred in 
small amounts even in disintegrated tissues, but the reduction of 
oxyhemoglobin must be regarded as due to the same constituent 
of tissue juices that reduces soluble Prussian blue or methylene 
blue; that constituent is not carbon dioxide, and we have shown 
that it behaves in more than one particular as does an enzyme. 
Further, the active reducer in tissues is not appreciably soluble in 
water; carbon dioxide is eminently so. Tissue reduction is the 
more rapid the higher the temperature, within the limits of 
enzymic action; but the higher the temperature the more rapid 
the evolution of carbon dioxide from tissue juices and, conse- 
quently, the less there is present to effect reduction of oxyhemo- 
globin. 

Barcroft has similarly shown^^ that traces of acid facilitate the 
dissociation of oxyhemoglobin. We feel convinced that the 
reductions we have studied have not been due to this cause, seeing 
that the older the juice the less vigorously did it reduce; whereas 
it ought to have reduced more vigorously with age, if the active 
agent in it had been acid formed by tissue autolysis. Further, 
traces of acid in our mixtures would have tended to form met- 
hemoglobin, a pigment we have never found in any mixture of 
active tissue juice and hemoglobin, however long they were left 
in contact. 

The factor of the dissociation of oxyhemoglobin with rise of 
temperature has not escaped us. We feel sure that it is not a 
potent factor in our experiments; for we have repeatedly observed 
that the oxyhemoglobin in contact with the boiled tissue juice in a 
control tube, remained unreduced at the end of many hours, when 
the active tube had been reduced in a few minutes. 

We do not wish to minimize the action of carbon dioxide, traces 
of lactic or other acid, and temperature in the dissociation of 
oxygen from oxyhemoglobin, but wc submit that there is both in 
cold-blooded and in warm-blooded animals an important enzymic 
factor in this phase of tissue respiration. We might remark that 

" J. Barcroft and L. Orbeli: Jour. Physiol., xli, p. 35o, 1910-11. 



196 Reduction of Hemoglobin by Tissue Reductase 

tissue juice reduction has been shown not to be due to alkaline 
salts,^' or to proteins,^* or to products of autolysis/^ or to bacterial 
decomposition,^^ or to catalase/^ as well as not due to dextrose 
or bile. , 

The differences in the behavior of the muscle juice of cat and of 
pigeon are noteworthy. The muscle juice of the cat, even when 
prepared before the heat had left it, was but a feeble reducer; 
whereas the juice of pigeon muscle was quite powerful. Cat 
muscle juice, moreover, both active and in boiled control, adsorbed 
much hemoglobin and so removed it from solution; the pigeon's 
did not. Our experiences with cat muscle are in agreement with 
the observations of some previous workers on the behavior of 
muscle in tissue respiration. Thus Sir Victor Horsley and A. B. 
Ilarris noticed^^ that when they injected methylene blue into 
living animals the voluntary muscles were very blue during 
activity, from which they concluded that in that condition oxi- 
dation was maximal arid was predominating over reduction. 
Further, the late C. A. Herter in his work on the reducing power 
of tissues^® found that Uver, lung, the suprarenals, and the gray 
matter of the central nervous system were all better reducers of 
methylene blue than were muscles or connective tissue. Herter 
also found that the postmortem decline of reducing power in 
rabbit muscle was very rapid. With the general tendencies of 
these conclusions we agree, and can confirm the observation on 
the rapid decline in activity of muscle after the death of the animal. 
This we would associate with the physicochemical immobiliza- 
tion which muscle undergoes as soon as the animal heat has left 
it. While unquestionably reduction of oxyhemoglobin in active 
muscle must l)c a constant occurrence of physiological importance, 
yet it is clear that oxidative processes necessarily predominate in 
such a tissue as muscle, in which heat is constantly being produced 
and energy of movement very frequently exhibited. 

*' Harris: Science Progress^ i, p. 730, 1907. 

»» Creighton: Tr. Nova Scotia Inst. .Sc, xiii, pt. ?, p. 61, 1911-12. 

*^ Harris: Biochem. Jour., v, p. 143, 1911. 

»^ Harris: ibid., v, p. 143, 1911. 

*" Harris and Creighton: Proc. Roy. Soc, Series B, Ixxxv, p. 486, 1912. 

** Victor Horsley and A. B. Harris: Brit. Med. Jour., ii, p. 205, 1895. 

>» C. A. Herter: Am. Jour. Physiol., xii, pp. 128, 457, 1904-05. 



D. F. Harris and H. J. M. Creighton 197 

The muscle juice of the pigeon, however, behaved quite differ- 
ently; we found it a very energetic reducer, although we have 
evidence that its power of reduction disappears rapidly after the 
death of the animal. While we should not place pigeon muscle 
at the head of a list of tissues arranged in the relative order of 
their reducing powers, as did Berns£ein,^^ yet we agree with the 
general results of this work in that we place the pectoralis major 
of the pigeon as second only to its liver in the intensity of reducing 
power. 

While we are not in a position at present to offer a satisfactory 
explanation of the difference in the reducing power between 
mammaUan and avian muscle juices, we can appreciate the 
physiological desirability of such a muscle as the bkd's pectoralis 
major being able with the utmost rapidity to obtain from its blood 
the maximal quantity of oxygen; in other words, to reduce that 
blood perfectly in as short a time as possible. 

The heart muscle juice (cat) was, as' compared with Uver juice 
(cat), a feeble reducer. Although there was some haziness 
observed between the two bands of oxyhemoglobin, reduction 
waa never complete as judged by our spectroscopic standard. 
Cat cardiac muscle, therefore, falls into the same category as 
regards reduction as the striated body muscle of the cat. Of 
course, the same remark applies to this substance; namely, that 
in it there are energy transformations of the same order as are 
going on in voluntary muscle. As judged by direct measure- 
ments, Barcroft^^ has shown that the heart muscle has only one- 
third the gaseous metaboUsm of the kidney. It may be further 
remarked that the hearts used were all from kittens; that is, 
immature animals in which it is well known to physiologists that 
tissue metabolism is less intense than in adult. The fact that 

" BernsteiD's work is thus described in Schafer's Te^t Book of Physiology, 
i, p. 782, Edinburgh and London, 1898: "Tissues placed in normal saline 
containing haemoglobin quickly reduce that substance, and in this respect 
muscle is the most effective. Bernstein found the following values for 
the rate ot reduction: muscle 100; liver 81.47; involuntary muscle 72.4; and 
the mucous membrane of the stomach 57.05" .... "This relative 
power of reduction holds good for tissues taken from frogs and from mam- 
mals." (Bernstein, Unteratich. a. d. physiol. Inst. d. Univ. HailCy i, p. 107, 
1888.) 

-» J. Barcroft and W. E. Dixon: Jour. Physiol., xxxv, p. 203, 1906-07. 



198 Reduction of Hemoglobin by Tissue Reductase 

the kidneys were also from kittens (immature animals) has to 
be remembered in connection with the relatively long time (fifty- 
seven minutes) taken by the renal reductase to effect complete 
reduction of the oxyhemoglobin in contact with it at 40°. It is 
well known that immature animals take a much longer time to 
drown than do adult; in terms of tissue respiration this may bv 
stated as due to a less intense oxidation of the tissues and, there- 
fore, a less rapid reduction of the blood supplying them. That 
there is some probabiUty in our surmise is supported by Vernon's^- 
work on adult cat tissues and on kitten's. Speaking of the in- 
tracellular ferment erepsin he writes: 'Tn the more active cat 
we find that the muscles of the adult contain twice as much fer- 
ment as those of a young kitten." But apart altogether from the 
age of the animal, the kidney of the cat does not rank as an active 
reducer. Judged by our criteria, the tissues of kidney are not in 
the same class as the liver as reducing agents. This finding would 
be in agreement with the statement by Barcroft: *^ Normally the 
blood leaving the kidney contains more than 60 per cent of its 
oxygen;" and again, *' Blood .... of the kidney is very 
arterial."^ 

As far as our experience has gone, liver juice is the most active 
reducing agent we have examined, and that of the pigeon the most 
active of all types of liver. This would seem to be in agreement 
with the conclusions^ of BatcUi and Stern working with tissue 
catalase. The late Prof. C. A. Herter, in his Harvey lecture 
(1906), in speaking of anaerobic conditions in bacteriological 
research said: ^*The reducing action of fresh liver has been suc- 
cessfully employed by Prof. Theobald Smith in rendering th(^ 
closed arm of the fermentation tube more strictly anaerobic." 

We have found reductase in the tissue of animals of four of the 
five great groups of the Vertebrata; namely, mammals, birds, 
amphibia, and fishes. The reductase of fish liver we found ex- 
ceedingly powerful both as regards oxyhemoglobin and soluble 
Prussian blue. It reduced cat blood as rapidly as it did fish blood. 
Its energy of reduction is very remarkable w^ien we consider that 
at the time of our examination of the liver of the fish, it had hvou 

-- H. M. Vernon: Jour. Physiol., xxxiii, p. 85, 1905-06. 

" J. Barcroft and M. Camis: Joiir. Physiol., xxxix, p. 134, 1909-10. 

'* M. F. Batelli and L. Stem: Arch. d. fisioL, ii, p. 471, 1905. 



D. F. Harris and H. J. M. Creighton 199 

ilead for alx)ut twenty-four hours. The physiological advantage 
to the fish to have a powerful reducer in its tissues is too obvious 
to be more than mentioned. Seeing that f'sh have at their dis- 
posal relatively such small amounts of oxygen, it is of the highest 
(*onsequence to them to be able to extract it from their blood with 
the greatest possible thoroughness. 

The biological significance of there being no specificity of 
interaction between the reductase of one animal and the oxy- 
hemoglobin of another, is that the oxyhemoglobin in the blood of 
A can be reduced by the reductase of B, A and B being in widely 
separated groups. Thus if human reductase were unable to 
reduce a foreign oxyhemoglobin, the perfusion into man of the 
blood of a lower animal would be absolutely useless in that the 
tissues could not cause that oxyhemoglobin to dissociate and so 
would be asphyxiated. It is, however, well known that when 
danger arises from the introduction of a foreign blood, it is not 
due to any difficulty about its being reduced, but to the liability 
of the human erythrocytes being hemolyzed by the blood of the 
lower animal. That the reductase of any animal can reduce the 
oxyhemoglobin of any other might be expected from the uni- 
formity of composition of hemoglobin throughout the animal 
kingdom. As far as reductase is concerned, there is no ^'specific- 
ity" of hemoglobin in the sense of Bradley and Sansum,^^ who 
used the anaphylactic reaction of the guinea pig to demonstrate it. 

We think that we may venture, in conclusion, to make a sugges- 
tion in connection with the hypothesis of the mechanism of tissue 
respiration which is given by Vernon and is based on Dakin's. 
N'emon writes :^^ 

The tissues contain a substance which can absorb oxygen from their sur- 
roundings to form an organic peroxide, and by the help of a peroxidase can 
transfer this oxygen to amino acid and carbohydrate molecules bound up 
in the tissues. . . . The organic peroxide, though it can still effect 
some oxidation, cannot of itself carry i1 to the final carbon dioxide stage. 

It seems to us that Vernon's ''substance which can absorb 
oxygen'* is none other than the tissue reductase whose behavior 
we have l>een studying. In the earlier part of this investigation.^" 

-'" H. C. Bradley and W. D. Sansum: this Journal, xviii, p. 497, 1014. 

-" Vernon: Jour. Physiol., xxxix, p. 182, 1909-10. 

" Harris and Creighton: Proc. Roy. Soc, Series B, Ixxxv, p. 487, 191?. 



200 Reduction of Hemoglobin by Tissue Reductase 

we pointed out that reductase could not only reduce compounds 
containing oxygen, such as methemoglobin and sodium nitrate, 
but such relatively stable substances containing no oxygen, as 
ferric chloride and soluble Prussian blue. In other words, the 
living tissues contain a substance with energetic reducing powers; 
we have brought forward a good deal of new evidence that this 
substance is an intracellular enzyme, which has been named 
reductase. It must be reckoned with in any hypothesis of tissue 
respiration. Starling^* does, indeed, to some extent do this: 
''There is no doubt that reducing substances are formed under 
normal circumstances in the tissues, as is shown by the meth- 
ylene blue experiment, and it is possible that such reducing sub- 
stances may aid in activating oxygen and in the induction of cer- 
tain oxidative processes." Our work enables us to be more 
definite in regard to these "* reducing substances;" it seems to 
warrant our asserting that the reductions effected by living 
tissues and by their juices in vitro are carried out by an endo- 
enzyme, already named reductase, some of whose properties we 
have investigated. It is not enough to describe the activity of 
an oxidase in this connection; our observations have shown that 
the tissues are energetic reducers on account of the presence in 
them of an enzyme, some of whose properties and behavior we 
have elucidated. If tissue respiration is completed by an oxidase, 
it is originated by reductase, 

V. SUMMARY. 

1. The reduction of oxyhemoglobin, at 40°, to the one-banded 
condition by the reductase contained in the liver and certain other 
organs of mammals, birds, amphibia, and fish has been inves- 
tigated spectroscopically. 

2. It has been found that the liver juice of all these classes of 
animals has the greatest reducing power, that of the pigeon being 
the most acrtive of all the animals studied; the muscle juice of the 
cat is but a feeble reducer; while, on the other hand, the muscle 
juice of the pigeon is only second in activity to its liver juice. 
Press juices of kidney, stomach, pancreas, heart, and cortex 
cerebri exhibit varying degrees of activity. 

*■ E. H. Starling: Principles of Human Physiology y London, 1912, p. 123. 



D. F. Harris and H. J. M. Creighton 201 

3. In addition to reducing oxyhemoglobin, the press juices of 
the organs studied reduce soluble Prussian blue to the leuco 
compound. 

4. It has been found that there is no specificity between reduc- 
tase and oxyhemoglobin, in that the reductase of any one animal 
can reduce the blood of any other. 

5. The influence of temperature on the rate of reduction of 
oxyhemoglobin by the reductase contained in cat Uver juice has 
been investigated, and it has been found that the rate of reduction 
increases with rise of temperature. Between 10° and 40° the 
temperature coefficient of the process of reduction is approxi- 
mately 2. Above 40° its value rapidly decreases with rise in 
temperature. Below 10° the rate of reduction is rapidly retarded, 
until below 0° inhibition of the enzyme is practically complete. 
Liver juice which has been kept below 0° for a time regains its 
original activity on raising its temperature. 

6. It has been shown that the activity of the reductase of cat 
liver juice decreases with the age of the juice, becoming prac- 
tically inactive at the end of about eight days. The curves 
obtained by plotting the reducing power of the liver juice against 
the age of the juice are logarithmetic. A kinetic investigation 
of the decay in the activity of reductase shows that the process 
follows the unimolecular law. 

7. The physiological bearing of the results of the investigation 
has been discussed. It is contended that reductase is the chief 
factor in causing dissociation of oxygen from oxyhemoglobin 
(reduction) in the tissues; carbon dioxide, traces of acid, and 
temperature being all of subsidiary importance. These obser- 
vations tend, as was hoped, to throw some light on the biochemical 
aspect of the inspiratory phase of internal respiration, the process 
whereby oxygen is removed from the oxyhemoglobin and from 
the lymph bathing the living cells, and applied in the nascent 
state to oxidisable substances in the cells. 

In conclusion, the bearing of the results of this investigation 
upon the process of tissue respiration has been discussed, and it 
has been shown that if tissue respiration is completed by an 
oxidase, it is originated by reductase. No statement as to the 
physicochemical rationale of tissue respiration is complete 
which does not recognize the existence and capabilities of reduc- 
tase. 



STUDIES IN CARBOHYDRATE METABOLISM. 

VIII. THE INFLUENCE OF HYDRAZINE ON THE UTILIZATION 

OF DEXTROSE. 

By frank p. underbill and ALBERT G. HOGAN.i 

{Front the Sheffield Laboratory of Physiological Chemistry ^ Yale University, 

New Haven.) 

(Received for publication, January?, 1915.) 

A long series of investigations has shown beyond question that 
the amount of dextrose in the blood varies little throughout 
normal life. No great differences are found even in the blood of 
(lifTerent species; and it may be said in general that the percentage 
of sugar in the blood is one of the body constants. This, of 
course, does not apply to pathological conditions, and for that 
n^ason natural and artificial abnormalities have been carefully 
observed. As knowledge has accumulated, a number of pro- 
cedures have been made available for producing variations in the 
blood picture, with respect to the sugar content. Most of these 
increase the amount of glucose, but by a few methods a condition 
of hypoglycemia can be obtained. One of the most effective 
agents in producing the latter condition is hydrazine, which was 
the medium of attack in this investigation. 

The earlier literature concerning the action of this compound is very 
scanty, though the toxicity of hydrazine has long been known. A complete 
account of the earlier pajycrs is found in the recent publications of Under- 
hill and his pupils, who have demonstrated the influence of hydrazine 
poisoning upon the sugar content of the blood. Undcrhill and Kleiner- 
showed in the first paper on the subject that the part it ion of the nitrogenous 
constituents in the urine was only slightly altered by the administration 
of moderate doses of hydrazine. The essential positive finding was that 
the poison exerts a selective action on the liver, as indicated by fatty 



*'The essential facts in this paper are taken from the dissertation pre- 
sented by Albert G. Hogan for the degree of Doctor of Philosophy, Yale 
I'niversity, 1914. 

- F. P. Underhill and I. S. Kleiner: this Journal, iv, p. 165, 11K)S. 

203 



204 Carbohydrate Metabolism 

infiltration. In a later research* a quantitative study of the blood was 
made, and it was discovered that one of the typical effects was a marked 
hypoglycemia. This condition was very constant in dogs; rabbits were 
more variable. A striking fact that came out in this study was the action 
of dextrose on dogs that had received hydrazine. The sugar when given 
subcutaneously in doses of 5 gm. per kilo promptly caused the death of 
the animal. This work was followed* by an investigation of the rela- 
tion of hydrazine to pancreatic diabetes. It was demonstrated at this 
time that the glycosuria which follows removal of the pancreas can ho 
temporarily inhibited by moderate doses of hydrazine. 

The facts that are now known concerning the toxicity of hydra- 
zine fail completely to afford any explanation of its action. If, 
however, the processes here involved were once made clear, it is 
probable that valuable information on the intermediary metal)- 
olism of carbohydrates would be available. In this connection 
recent work of Frank and Isaac* deserves consideration. Thcv 
poisoned animals with phosphorus, which in some respects re- 
sembles hydrazine in its action. It causes hypoglycemia and the 
rapid disappearance of glycogen. In both cases, also, the liver 
shows a characteristic yellow color, presumably due to fatty 
infiltration. As was pointed out in the paper of Underbill and 
Kleiner, however, there are some essential differences in the action 
of these two poisons. Hydrazine limits its action to the cells of 
the liver and affects only the cytoplasm; the cells first attacked 
are in the center of the lobules. In phosphorus poisoning the 
periphery of the liver seems to be injured first. It attacks the 
cell nuclei and does not limit its action to that organ. Frank 
and Isaac in their study of phosphorus poisoning devoted con- 
siderable attention to the sugar of the blood. They administered 
dextrose to rabbits under different conditions, and observed the 
time necessary for the blood sugar content to become normal. 
The sugar was administered through a stomach sound in doses 
of about 10 grams per kilo. When given to a normal rabbit the 
blood sugar content increased to about 0.3 per cent and became 
normal again in about four hours. When the dextrose was given 

" F. P. IJnderhill: this Journal, x, p. 159, 1911-12. 
* F. P. Underbill and M. S. Fine: this Journal, x, p. 271, 1911-12. 
See also Underbill: this Journal, xvii, pp. 293 and 295, 1914. F. P. 
Underbill and A. L. Prince: this Journal, xvii, p. 299, 1914. 

5 E. Frank and S. Isaac: Arch. f. expcr. Path. u. PharmakoL, Ixiv, p. 
274, 1911. 



F. P. Underbill and A. G. Hogan 205 

shortly after the phosphorus had been administered, the course of 
the blood sugar closely followed the normal curve. If the interval 
between the introduction of phosphorus and dextrose was in- 
creased to twenty-four hours, however, then the abnormally high 
sugar content was reduced very slowly, being delayed ten hours 
in some cases. Since phosphorus produces results so similar to 
hydrazine in some respects, and so unlike it in others, it seemed 
desirable to make similar trials with the latter drug. 

EXPERIMENTAL. 

Plan. The plan adopted was to administer hydrazine to an 
animal, and at the time when its effects were most evident to 
introduce dextrose subcutaneously. The result of such a sub- 
cutaneous injection both in a normal and in a poisoned animal js 
to raise the blood sugar content. Underbill showed that hydra- 
zine may cause marked hypoglycemia in rabbits, a finding that 
was confirmed in this investigation. If this reduction is due to 
some increase in the glycolytic processes, it would seem prob- 
able that dextrose administered in lar^e quantities would be 
promptly disposed of, more quickly than by a normal rabbit. 

Methods. The experimental animals used were rabbits. The 
general plan was to administer the maximum non-fatal dose of 
hydrazine, and then to make subcutaneous injections of dextrose 
at the time when the sugar content of the blood was at its min- 
imum. The method of determining the blood sugar was that 
devised by Forschbach and Scverin.® Before attempting the 
above experiments it seemed desirable to determine the optimum 
experimental conditions; as, for example, dosage of hydrazine 
and the amount of dextrose to be injected. Other factors in- 
volved were the effect of starvation, as animals will not eat after 
the administration of hydrazine, and the normal time required 
for the utilization of the amount of dextrose to be introduced. 

The optimum conditions for the production of hydrazine hypogly- 
cemia in rabbits. 

The first point established was the most suitable dosage of 
hydrazine for the production of hypoglycemia, together with the 

•Forschbach and Sevcrin: Arch. f. exper. Path. u. Pharmakol., Ixviii, 
p. 341, 1912. 



THK JOURXAL OP DIOLOOICAL CH£UI8TRY, VOL. XX, KG. 3 



2o6 Carbohydrate Metabolism 

time required for it to exert its action. Hydrazine sulphate 
(Kahlbaum's) in 2.5 per cent solution was administered subcu- 
taneously; the dose varied from 60 to 85 mgm. per kilo of body 
weight. At various intervals the per cent of dextrose in the blood 
was determined, in order to find what dosage was most effective, 
and how much time elapsed after its administration before the 
blood sugar reached its lowest value. The general method for 
obtaining blood was to fasten the animal on a rabbit board in a 
horizontal position, and make a transverse incision in the marginal 
ear vein. The blood was allowed to fall directly into the weighing 
bottle; about twenty or thirty drops are required. If the blood 
flows too slowly it can be hastened by warming the ear with an 
electric light bulb. Table I contains the experimental data. 

The influence of inanition on the sugar content of the blood. 

Inasmuch as rabbits refuse food when under the influence of 
hydrazine, it became essential to determine the influence of 
starvation alone upon the blood sugar content. The general 
plan adopted was similar to that given above. The animals 
received no food, and the blood sugar was determined daily 
through a period of approximately one week. The results are 
given in Table II. 

The results in this table indicate that starvation does not influ- 
ence the per cent of sugar in the blood. 

The rate of disappearance from the blood of glucose subciUaneously 

injected into starving rabbits. 

In order to formulate a satisfactory working basis, a series of 
determinations in normal animals was made to show how rapidly 
a subcutaneous administration of glucose would disappear. This 
was done by injecting various amounts of sugar and examining 
the blood at frequent intervals. Food was withheld from the 
rabbits for a period of two days before injecting the dextrose. 
For the results sec Table III. 

These experiments demonstrate that if 3 to 5 grams of dextrose 
are administered subcutaneously to normal rabbits the percentage 
of blood sugar usually regains the normal within a period of three 
to four hours. 



H 






K 
a 



00 



CO • tA 

^ ,-4 00 



8 





kC 




1^ 


• 

1^ 


s 




t^ 




''J* 


■ 


^ 


1^ 


f-4 


t^ 



00 



C^4 



OS 



:: S 8 S iS 



CI 



08 



eo 



CI 



»o 









CO 



CI 



2 3 8 8 5 



8 3 






•S i 





^ -5 
a o 

••• 

N 

Q 






c^ 






C« 



oc 



00 


• 


t^ 


• 

o 






CO 

eo i-< 

o 



52 









s 















■s 




« 




p 




8 


OS 


• 


a> 


o 


Q 


s 


eS 


• 


o 


o 


Q 



s 



s 



s 



2 8 



& 2 



><» 



« 



8 



CD 

^ cS 



s 



CO 



2 S 



8 8 



o C^ 



8 8 



8 S 



S g S 



8 



Ol 



CO 



c^ 



Ol 






8 8 SS3 



• • 



Sep O CO C4 

O f-4 f-4 i>iH 



• • 



CO 



»i3 

< 

O 

o 
2 



eo 
O 



H 



X 
K 
C 

Z 

Q 

b 

C 

o 

a 



« 2 

< * 

S g 

D « 

O o 

o pa „ 

J t> S 



as 



>* 

X 

o 

8 

c S a 

H £ N 



-< 
o 

P 
eo 

O 

H H 

H 
0* 



06 
H 



a 
o 
o 



< 



s 






207 



208 



Carbohydrate Metabolism 



The rate of disappearance fram the blood of glucose suhcuianeoushj 

injected, into hydrazinized rabbits. 

Finally determinations were made to find how rapidly dextrose 
disappeared following administrations of hydrazine. In this 
series the normal percentage of blood sugar was obtained, and 
then hydrazine was given. Two days later the blood sugar con- 
tent was again estimated. The dextrose was then administered 

TABLE II. 

The effect of inanition on the per cent of sugar in the blood. 



i 

1 




PBB CAMT OF SUQAB IN THE BLOOD 




NO. OF 
ANIMAL 


Normal 


' 


AftOT a starvation period ot 




1 


2 dya. 


1 


Sdyft. 
0.15 


4 dys. 


17 


0.12 


0.12 


1 

1 


0.12 


18 


0.12 


1 0.12 




0.11 


0.12 


19 


0.10 


0.10 




0.14 


0.11 



TABLE III. 
The content of blood sugar at intervals after subciUaneous injections of 

dextrose. 



NO. OF ANIMAL 

WEIGHT IN KILOS , 

GM. OF DEXTROSE PER KILO 

INJECTED 

IN 30 PER CENT SOLUTION 
IN 15 PER CENT SOLUTION 

PER CENT OF BLOOD SUGAR 



A*TER INJECTING DEXTROSE . . 
" inn 


tt 


» — --.... 

i " 


u 


A « 

A •••■■•■•••■•*■>• 


it 


1 " 


it 


U " 


it 


2 " 


it 


2i " 


ti 


3 " 


a 


3-5 '* 


it 


J 

4 '* 


11 


4i " 


a 


5 ** 


<t 


(» ** 



! 20 


21 


22 


23 


27 


. 28 


29 


1.7 


1.2 


1.8 


1.4 


1.9 


1.6 


1.8 


5 


5 


5 


o 

• 


5 


1.5 


3.0 


0.12 


0.17 
0.33 


0.11 
0.30 


0.13 
0.28 


0.18 


0.09 




0.22 


0.34 


0.28 


0.36 


0.31 


0.19 


0.33 


0.27 


0.33 


0.31 


0.35 








0.30 


0.33 


0.35 ■ 


0.36 


0.46 


0.09 


0.33 


0.31 


0.25 


0.24 


o.as 


0.53 


O.OS 


0.28 


0.25 


0.18 


0.20 


0.31 


0.48 


O.as 


0.28 


0.30 


0.15 


0.20 


0.25 


0.47 


0.12 


0.23 


0.25 


0.07 


0.10 


0.20 


0.40 


0.09 


.0.18 




0.09 1 


0.14 ' 


0.18 


0.40 




0.1.') 




0.13 , 


0.15 


0.22 


0.39 




0.14 




0.16 ' 


0.15 
0.15 


0.20 
0.24 


0.3G 
0.30 






0.13 








0.30 







F. P. Underbill and A. G. Hogan 



209 



!_ ^ 00 QQ C^4 

OC0C*_^^^ CI CI 

1 CO • QC CO 

— • 00 000 

CO • OC CO ^ • • 

^ 00 000 

^00 1-^ p o 

^ 00 000 

?? '^ g * '^' -^ -• ^^ 

^ 00 00 

? N "^ ».'3 * _, rH O CO S rH 

•*j ' CO • t"^ CO • " 

H 1-1 00 000 

. » CO t>. »>. o o 

'^ "CO • t>- CO '^ • • 

^ 1-H 00 o o o . 

^ - .- — ____ 

■^ ^ <M O »Q CI CO t>- "^ OC i-^ 05 00 

5: coo>'^'^ ^,1-11-1 c5 cococicow>"^co?5 

M CO I- CO -1 . . 

i;^ ^ 00 O 00000000 

,_ O O >?: ».0 r>i '^ O CI CO rt< CO -^ CO 

^ C« • l^ 'M . . . 

,-1 CI 00 o 00000 

? 'f -^ oi »5 -^ CI CI CI t* 00 o t> r^ rH o 

iftCOO* -yK rH d^lCO'^COCOCOClClCIWCICOCO 

■ CI • i^ CO " 

^^ 1-1 o 00000000000000 

a^ ^O COOQOOOCOr^COiOCIiO-^Q Q 

^ - ^ '^ »fi * ^, ^rH CO'^iHOOCOCO'^OOOiOCJ c5 

;C p* CI • t>. O - « . . 

^? CI 00 0000000000000 o 



'■C 



►I 






2: 



•^ 






> w 



•-> 




X. 


*"* 


"* 


i< 


^» 


b; 


■^ 


X 



" X ^ 












Z 



c" a 
y. ^ 



Q 

S 






X 

c a 



X 



O 






a c 
a 

u 

Q 



u 

21 CC 

i; o 



>5 



-< 

p 

CC 

Q 
O 
O 

n 



0: 
o 

a 



c 

Si 
H 
X 

is 



* HC« rt"* 



»<.« H« "*« "^w 

CI CI CO CO rt< Tj« 1^ O O 1^ 



c r: 

r c 



n c 

a a 






2IO Carbohydrate Metabolism 

subcutaneously, and the amount of sugar in the circulation was 
determined at varying intervals. The results are given in 
Table IV. 

A comparison of Tables III and IV indicates that hydrazine 
markedly retards the utilization of dextrose after subcutaneous 
injections of the sugar. 

DISCUSSION. 

A survey of the data tabulated indicates that, when introduced 
into rabbits, 80 mgm. of hydrazine sulphate per kilo is probably 
the largest dose that can be given without lethal effects. Sev- 
enty-five mgm. is a safer amount, and perhaps just as effective. 
It is quite evident, however, that the reaction of the rabbit to 
hydrazine is very inconstant. While the blood sugar of dogs is 
reduced to minimal values with a dosage of 50 mgm. per- kilo, it is 
frequently impossible to obtain that result when using rabbits, 
whatever amount is given. As might be expected, the amount 
of sugar in the blood during starvation does not vary from normal. 
Starvation alone, therefore, plays no significant part in the action 
of hydrazine in respect to the content of blood sugar. 

When dextrose is injected subcutaneously into a normal rabbit , 
there is considerable variation in the rapidity with which it dis- 
appears from the circulation. The concentration of the solution 
probably has some effect on the time required for absorption. A 
comparison of Tables III and IV shows, however, that if dextrose* 
is subcutaneously administered during hydrazine poisoning, the 
resumption of a normal blood sugar content is markedly delayed 

SUMMARY. 

Hydrazine causes hypoglycemia in rabbits as in dogs, but not 
so consistently. 

Starvation causes no change in the content of blood sugar. 

When dextrose is administered two days aiter the administra- 
tion of hydrazine, its utilization as indicated by the blood sugar 
content is markedly retarded. This is true whether the amount 
of sugar in the blood was previously low or not. 

These facts offer no explanation for the observed diminution 
of dextrose in the blood after the administration of hydrazine. 



STUDIES IN CARBOHYDRATE METABOLISM. 

IX. THE INFLUENCE OF HYDRAZINE ON THE GLYOXALASE 

ACTIVITY OF THE LIVER. 

By frank p. UNDERBILL and ALBERT G. HOGAN.* 

(Prom the Sheffield Laboratory of Physiological Chemistry, Yale University, 

New Haven.) 

(Received for publication, January 7, 1915.) 

Modem theories of metabolism assign vast importance to 
enzyme action. These hypotheses, though completely in accord 
with the facts as determined in vitro, should be extended to proc- 
esses in the bpdy with great caution. In accordance with our 
present ideas, however, it seems quite possible that the peculiar 
effect that hydrazine^ produces on the sugar content of the blood 
may be due to some interference with the essential enzymatic 
activities. In this connection three possibilities suggest them- 
selves. The enzjmae or enzymes concerned may be accelerated, 
retarded, or entirely destroyed. 

EXPERIMENTAL. 

Problem. In view of these possibilities the plan adopted for 
this investigation was to make a series of observations on the 
influence of hydrazine on the glyoxalase activity of the liver. As 
Dakin and Dudley,^ who discovered glyoxalase, have pointed 
out, it is very active in liver extracts. Furthermore, it is prob- 
ably directly concerned in the intermediary metabolism of car- 

^ The essential facts of this paper are taken from the dissertation pre- 
sented by Albert G. Hogan, for the degree of Doctor of Philosophy, Yale 
University, 1914. 

'Compare previous paper: this /ourna/, xx, p. 203, 1915. 

» H. D. Dudley and H. W. Dakin: this Journal, xiv, pp. 165, 423, 555, 
1913; also xv, pp. 127 and 463, 1913. 

211 



212 Carbohydrate Metabolism 

bohydrates. The reaction involved is reversible, and proceeds 
as follows: 

CH3 CO CHO + H2O = CHa CHOH COOH. 

The results of Dakin and Dudley were confirmed and extended 
by Levene and Meyer/ They showed that leucocytes or kidney 
tissue if allowed to act on methyl glyoxal will convert it to lactic 
acid. Inasmuch as hexoses are also converted to lactic acid in 
this way, it is rendered even more probable that methyl glyoxal 
is an intermediate stage in carbohydrate metabolism. If, there- 
fore, glyoxalase is concerned in this process^ it would seem espe- 
cially adapted to the problem here involved. 

Methods. Eecause of its ready response to hydrazine poison- 
ing, the dog was chosen as the experimental animal. The dosage 
employed was 50 mgm. of hydrazine sulphate (Kahlbaum's) per 
kilo, and was administered subcutaneously in 2.5 per cent solu- 
tion. The maximum effect of the compound is usually attained 
within forty-eight hours; so on the second day after administering 
the drug the dog was killed. The procedure following was essen- 
tially that devised by Dakin and Dudley. The liver was finely 
ground and a 20 per cent extract prepared with distilled water. 
This was kept in a thermostat at a temperature of 35° or 37°C. 
for one hour, and then strained through several layers of muslin 
cloth. All manipulations were aseptic as far as possible, and no 
preservatives were used. Putrefaction rarely occurred and seemed 
to have no appreciable effect on the amount of mandelic acid 
present. Dakin found that phenyl glyoxal has certain advan- 
tages over the methyl compound for experimental purposes, so it^ 
was used in this investigation. It undergoes the same type of 
reaction, and the mandelic acid formed has a much higher specific 
rotation than lactic acid. A quantitative estimation, therefore, 
is much more accurate. 

The proportions used were 0.2 of a gram of phenyl gl^'oxal, 
dissolved in 5 cc. of distilled water, and 50 cc. of the tissue extract. 
The phenyl glyoxal solution wsis filtered before use. Freshly 

* P. A. Levene and G. M. Meyer: this Journal^ xiv, p. 551, 1913. 
» We are greatly indebted to Dr. Dakin for 8uppl3ring us with a large 
quantity of phen\l glyoxal. 



F. P. Underbill and A. G. Hogan 213 

precipitated calcium carbonate was added to the digestion mix- 
ture to preserve neutrality. At the end of the incubation period 
the mixture was removed from the thermostat and boiled. In 
order to determine the relative amount of mandelic acid, the 
procedure was as follows: 30 grams of ammonium sulphate were 
added to the solution and the mixture was heated on the water 
bath for three minutes. This was cooled and acidified with 
syrupy phosphoric acid. The precipitate was filtered off and 
washed with a solution of ammonium sulphate. Three cc. more 
of the phosphoric acid were added, and the mandelic acid was 
extracted by shaking out with ether; four extractions are sufficient. 
The ether extract was washed carefully with small quantities of 
water to remove traces of phosphoric acid, and then evaporated. 
The residuum was taken up in a Uttle water, and filtered into a 
200 mm. polariscope tube, with a capacity of about 17 cc. The 
tube was almost completely filled with wash water from the fun- 
nel, and after closing it was inverted repeatedly until the contents 
were thoroughly mixed. After determining the rotation the 
contents of the tube were transferred to a beaker, and titrated 
against -ft sodium hydroxide, with phenolphthalein as an indicator. 
Since the speed of the reaction was the point of greatest interest, 
a number of these estimations were made for each animal. These 
were incubated for different periods ranging from thirty minutes 
to nine hours in most cases." One was boiled before adding the 
phenyl glyoxal, and served as a control. In order to siscertain 
whether the hydrazine was producing its typical effects, deter- 
minations of the blood sugar were made. With one exception the 
amount was found to be very low. Dakin has demonstrated that 
pancreas extract inhibits the enzyme glyoxalase, so an experiment 
was devised to determine whether the pancreas of a hydrazinized 
animal also exercises this inhibition. Fifty cc. of a 20 per cent 
pancreas extract were added to 50 cc. of a 20 per cent liver extract 
of the same animal, which had received hydrazine. A similar 
mixture was made with a boiled pancreas extract to serve as a 
control. In a few cases the effect of using only a 10 per cent 
tissue extract was tried. The results of the various trials are 
given in the tables. 



214 



Carbohydrate Metabolism 



TABLE I. 
Normal liver. 



RATE or QLTOXALABB ACTIVITT 

20 per cent extract 



M 
g 



a 

i 

1| 

2! 
31 

6l 

9 

12 

2<y 

I 
I 

1 

2 

3 

6' 

9 



ROTATIO^f 



■5.0*V. 
-4.8* V. 
•4.2*V. 
-4.5'V. 
■7.5*V. 



-2.2* V.' -1.1*V. 
-4.3* V. -1.9'V.I 
-5.4«V.! -2.6* VJ 

-4.4«vJ -3 rv. 



-1.7* V.i -1.3'V. 
-2.8* V.i -2.4* V. 
-5.5* V.' -3.7* \. 
-5.8'V. -5.rv. 



-0.4* V. 



-5.4* v.: -5.3'V. -6.0* V. 



10 percent 
, -2.5'V. 

-2.6* V. 

-3.3»V. 

-3.8* V. 

20 per cent 





'Control. 




I 



-4.1* V. 



extract 
-0.8* V. 
-1.3* V. 
-1.3* V. 
-2.0* V. 
extract 




10.51 
9.95 
9.75 
9.90 

13.80 



No. of animal 



I 



|(boiled); 

! 8.7 • 



ACIDITY 

CO. ^ NaOH 



2 



3 



8.00 


1.05 


9.ft5 


1.8 


10.16 


2.9 


14.55 


2.8 



10.1 



2.82 



4.7 



2.5 
3.5 
6.6 
7.4 

7.2 



2.4 
3.1 
4.0 
3.8 

0.1 



2.2 
3 4 
5 6 
7.1 



2.3 
2.9 
2.9 
3.9 

0.45 



TABLE II. 

Hydrazinized liver. 



1 

2 
3 
9 



RATB or OLTOXALABE ACTIVITT 

20 per cent extract 



HRS. INCUBATED 






" 6 ' 


\ 


-1.2'V. 


I 


-2.9«V.i 


2 


; -4.4* V. 


3 


-4.8»V. 


6 




12 


-5.6* v.: 


20 


-4.8»V. 


>)ntrol (9 hrs.) 






ROTATION 



No. of animal 



7 8 

-i.4*V.; -1.2«V. 

-2.3* v.! -2.6*V. 

-4.9*\'.- -4.5* V. 

-4.4* V.i -5.7*V. 

-5.3' V.l -5.5* V. 

I 



9 

-1.2* V. 
-2.5'V. 
-4.4»V.i 

-4.8* v.; 

-4.3* V. 



10 



-0.2*V. 
per cent, 

-i.rv. 

-2.0* V. 



I 
extract i 
-0.7«V.' 
-1.4»V. 
-2.0*' V. 
-3.0'V.; 



4.5 
5.2 
6.6 
8.1 

10.0 
7.7 
1.1 



ACIDITr 



7 


1.9 


3.1 


6.5 


6.3 


6.7 



0.6 



8 

1.8 
3.9 

5.8 
7.2 
8.1 



1.2 



1.6 



3.0 



H 

2.2 
3.6 
5.4 
5.7 
5 S 



6 



1 
2 



4 



F. P. Underbill and A. G. Hogan 



215 



NO. OF I 
ANIMAL , 



1 

2 

3 
4 
5 



6 
7 

8 
9 



TABLE m. 

Comparison of results. 
Normal dogs. 

BOTATIOK OB8BBVBD AITKR IKCUBATIKQ 



ihr. 



2.2^ V. 

i.rv. 

1.7^ V. 
1.3^ V. 



1.2«V. 
1.4^ V. 
1.2'»V. 
1.2'»V. 



Ihr. 



4.3^ V. 
1.9^ V. 
2.8^ V. 
2.4^ V. 



2hrs. 



-5.4*' V. 
-2.6** V. 
-3.7*' V. 
-3.7* V. 



Hydrazinized dogs. 



-2.9** V. 
-2.3* V. 
-2.6* V. 
-2.5* V. 



-4.4* V. 
-4.9* V. 
-4.5* V. 
-4.4* V. 



3 hrs. 

-5.0* V. 
■4.4* V. 
-3.1* V. 
-5.8* V. 
■5.1* V. 



4.8* V. 
4.4* V. 
5.7* V. 

4.8* V. 



9 hrs. 



4.2* V. 
5.4* V. 
5.3* V. 
6.0* V. 
4.1* V. 



-5.3* V. 
-5.5* V. 
-4.3* V. 



An extract of the pancreas, in the case of Dog 8, was added to 
the liver extract, and was found to inhibit completely glyoxalase. 
The unboiled pancreatic extract had no effect whatever. 



SUMMARY. 

A comparison of results indicates that the glyoxalase activity 
of the liver is not markedly altered by the action of hydrazine. 

It is, therefore, evident that the results of this investigation 
offer no explanation for the disappearance of glycogen from the 
liver and the diminished blood sugar content observed after 
hydrazine administration. 



A COLORIMETMC METHOD FOR THE ESTIMATION 

OF AMINO-ACH) a-NITROGEN. 

By victor JOHN HARDING and REGINALD M. MacLEAX. 
{From the Biochemical Laboratory, McGill Univerntyj Montreal.) 

(Received for publication, January 14, 1915.) 

Two methods are at present in use for the dotormination of 
amino-acid nitrogen. The first, that of Sorensen,^ is a rapid, 
easy method, but is neither very delicate nor accurate. The 
second, that of Van Slyke,- is the method now usually accepted 
for the determination of nitrogen in amino groups. In its latest 
development, the micro chemical form,' the method will estimate 
0.5 mgm. of nitrogen with an accuracy of 1 per cent, and its 
application to the determination of amino-acid nitrogen in blood 
and tissues has yielded in its author's hands very valuable and 
interesting results. It was felt, however, that a third and inde- 
pendent method for the estimation of the a-nitrogen of the amino- 
acids, especially if the sensitiveness of the reaction could be 
increased beyond that of the Van Slyke method without loss of 
accuracy, would not only be valuable but necessary as the time 
approached for a study of the chemistry of the single cell. 

It soon became apparent that such a hope could be fulfilled 
only by the quantitative application of a color reaction of amino- 
acids, and of all such the most probable seemed the reaction with 
triketohydrindene hydrate. 

This reaction was discovered by Ruhemann,^ who found that 
all acids containing a free amino group in the a position reacted 
with triketohydrindene hydrate with the production of an in- 
tense blue color. /3-, 7-, and 5-amino-acids only gave small amounts 
of color, and a-amino-acids substituted on the amino or carboxyl 

* S. P. L. Sorcnsen: Biochem. Ztschr., vii, p. 45, 1908. 

« D. D. Van Slyke: this Journal, ix, p. 185, 1911 ; xii, p. 275, 1912. 
» Van Slyke: ibid., xvi, p. 121, 1913-14. 

* S. Ruhemann: Tr. Chcm. Soc, xcvii, pt. ii, p. 2025, 1910. 

217 



2i8 Estimation of Amino-Acid a-Nitrogen 

group did not react at all. This discovery was confirmed and 
extended by Abderhalden, who applied it to the detection of 
pregnancy and cancer, triketohydrindene hydrate now being a 
commercial product under the name of "ninhydrin." The 
sensitiveness of the reaction is unquestioned; for, according to 
Abderhalden and Schmidt,* it will detect one part of glycine in 
65,000 parts of water, though the other amino-acids will not react 
in quite so dilute a solution. 

The constitution of the blue coloring matter was investigated 
by Ruhemann,* who isolated from the interaction of alanine 
and triketohydrindene hydrate a body which was found to be 
identical with the ammonium salt of diketohydrindylidene'diketO' 
hydrindamine, 

(0NH4) 

Thus the chemistry of the coloring matter was well known and 
the problem became the determination of the conditions under 
which the grouping 

RCH (NH2) CO2H 

common to all a-amino^cids would react quantitatively with 
triketohydrindene hydrate to produce the blue colored ammon- 
ium salt of diketohydrindylidene-diketohydrindamine, which could 
then be compared colorimetrically with a definite amount of 
coloring matter as standard. 

The reaction has already been made use of in a roughly quan- 
titative way by Abderhalden and Lamp^^ in studying the fate of 
amino-acids during absorption, but they made no attempt to 
place their results on a strictly quantitative basis. Herzfeld,' 
however, devised a method for the estimation of the group 

*E. Abderhalden and H. Schmidt: Ztschr. /. physiol, Chem., Ixxxv, p. 
143, 1913. 

• Ruhemann: Tr, Chcm. Soc, xcix, p. 1486, 1911. 

^ E. Abderhalden and A. E. Lamp6: Ztschr. f. physiol. Chem.y Ixxxi, p. 
473, 1912. 

« E. Herzfeld: Biochem. Ztschr,, lix, p. 249, 1914. 



V. J. Harding and R.*M. MacLean 219 



yNHa 
-CH<r 

Ndooh 

using the ninhydrin reaction as a basis. The method was to 
evaporate the amino-acid and excess of triketohydrindene hydrate 
to dryness on a water bath, dissolve the purple colored residue in 
a little alcohol with a drop or two of anunonium hydroxide, make 
up to a known volume, and determine the amount of coloring 
matter by measuring its extinction coefficient in a spectrophoto- 
meter. The present authors have repeated this method of pre- 
paring the coloring matter in a quantitative way, estimating it, 
however, by a Duboscq colorimeter instead of measuring the 
extinction coefficient in a spectrophotometer. The former 
method is much more rapid, and at present is the only one which 
could be successfully applied in hospital laboratories to the study 
of amino-acid excretion in pathological conditions. Working in 
this way, the results were very unsatisfactory. It was found 
by heating varying amounts of an amino-acid with an excess of 
ninhydrin that the different amounts of the amino-acid could be 
estimated with moderate accuracy, using a fixed amount of the 
same amino-acid as standard. The following figures illustrate 
this point. 



AMINO-AaD 



Glycine. 



Alanine. 



Aapartic acid 



Glutaminic acid 



OC. OF 0.1 PBB CXMT 
SOLUTION 

1.0 Standard 

1.5 

2.0 

1.0 Standard 

2.0 

4.0 

1.0 Standard 

2.0 

3.0 

1.0 Standard 

1.5 

3.0 



COLORIIIBTXR RBADINO 



Found: 



Calculated: 



Set at 2.00 cm. 
1.51 1.33 

1.00 1.00 

Set at 2.00 cm. 
0.99 1.00 

0.48 0.55 

Set at 3.00 cm. 
1.51 ! 1.50 

0.95 1.00 

Set at 2.00 cm. 
1.33 1.33 

0.69 0.66 



Thus it will be seen that varying amounts of glycine can be 
estimated by using the color produced by a known amount of 



220 Estimation of Amino-Acid a-Nitrogen 

glycine as standard. The same is true for alanine, and aspartic 
and glutaminic acids. 

When, however, we attempted to estimate alanine, or aspartic 
or glutaminic acids, using glycine as a standard, the method gave 
totally erroneous results. 



AMINO-ACID 


1 


Nt PER CC. 




Found: 

1 




Calculated 




mgm. 




mgm. 


Alanine 


0.016 




0.036 


Glutaminic acid 


' 0.152 


0.096 

i 



In the case of aspartic acid the color produced by the reaction 
was of such a pronounced reddish shade that it was found impos- 
sible to match it against the bluish violet color of the standard. 

Thus it will be seen that this method of estimating amino-acid 
nitrogen fails completely. 

Reaction of amino-acids and triketohydrindene hydrate in 

slight excess. 

In our earlier experiments on this subject, attempts were made 
to study the reaction between various amino-acids and the tri- 
ketone, with strictly molecular amounts of the two substances, but, 
owing to causes at present unknown, these experiments failed to 
give any concordant results. Neither of us, working alone or 
together, could be certain of obtaining any two series of experi- 
ments which agreed within 5 per cent. What we desired to do 
was to heat a known amount of an amino-acid, usually glycine oi- 
alanine, with the molecular equivalent of triketohydrindene hy- 
drate and deteraiine the time at which the maximum development 
of the blue coloring matter took place. This could be accom- 
plished by immersing a series of test-tubes containing the equiva- 
lent amounts in a boiling constant-level water bath, removing them 
one by one at stated intervals (usually five minutes), diluting them 
to a known volume (100 cc.) with distilled water, using the color 
produced in the test-tube first removed as standard, and comparing 
the others against that in a Duboscq colorimeter. At first the com- 
parisons were made by daylight, but it was soon found that mucli 



V. J. Harding and R. M. MacLean 221 

more consistent readings were obtained by the use of a 25 watt 
tungsten lamp as a source of illumination in a dark room. The lamp 
was placed in a conical, semi-opaque shade, the outer end of which 
was covered with a sheet of tissue paper to diffuse the light; and 
the whole was placed about a foot away from the colorimeter. 
In this way a strongly and evenly illuminated field was obtained 
on the white reflector of the colorimeter only, and thus the eye was 
free from any other disturbing sources of light during the deter- 
minations. A small electric bulb on a convenient switch enabled 
one to take the vernier readings. In artificial light the bluish 
colored solution of the coloring matter changes to a red violet, 
resembling very much the appearance of a dilute solution of 
potassium permanganate when viewed in daylight. 

Even under these conditions, which greatly improved our 
ability to make concordant readings, the results continued to l>e 
very irregular and capricious, when using equimolecular amounts 
of the two reagents. The presence of traces of impurity, the 
rapid fading of the coloring matter, and the sensitiveness of the 
reaction to heat, probably were all factors, not under these con- 
ditions properly controlled, which contributed to the failure of 
our experiments from a quantitative point of view. However, 
under more stable and more accurately defined conditions, it is 
hoped at a later date to return to the study of the reaction in 
equimolecular solution and to endeavor to obtain from it some in- 
sight Into the mechanism of what is without doubt a complicated 
series of reactions. 

It was soon found, however, that a slight excess of triketo- 
hydrindene hydrate enabled us to approach a solution x)f our 
problem of the time of the maximum development of the coloring 
matter in dilute solution. The same method of procedure was 
adopted as that previously described. The coloring matter in the 
test-tube first removed was used as standard and the others were 
read against it. The series of curves in Figure 1 shows the relative 
amounts of coloring matter plotted against the time, each amino- 
acid being taken separately. 

It will be noticed that the time of the development of the 
maximum amount of color is different for each amino-acid, thus 
precluding the use of this technique for a method of estimating 
amino-acid a-nitrogen in a mixture of amino-acids. No attempt 

THE JOUBNAL OF BIOLOGICAL CHBMZ8TBT, VOL. XX, NO. 3 



333 Estimation of Amino-Acid a-Nitrogen 

was made to compare one amino-acid against another, as such a 
series of experiments would have served no useful purpose at 
that time. Moreover, the actual amounts of color produced in 
the cases of aspartic and gluta,minic acids and asparagine were so 
small that the reaction liquid wa^ only diluted to 50 cc. instead of 
100 cc. Also the color produced by aspartic acid and asparagine 
was of a pronounced yellowish tint and could not be compared 
with the reddish violet given by alanine or glycine. 





' 






















































:'•!> 


«n^ 


h' 




-M- 


- 












^ 




















c 






^ 




















' n 




E 






y 




1 
















L-H 


w- 


t 








g 


^--;;j^ 


, 


— i 


-A 







-f^ 




^ 


: 


/^ 




^:^ 


^. 










) 




= /?J<? 










« 


" 


ri^ 












:ifi^^ 


V~ 




-V- 




_ 


_ 


^—^r^,. 








; 
















- 


- 


I 






1 


4- 






_ 












L 




1 






L- 


— i— 


L 


_ 




4- 




U 



Fig. 1. 



Reaclion of amino-acids and trikelvkydrindetie kijdrale in 
targe excess. 

In our next series of experiments wo determined the time of the 
maximum development of color between the different amino-acids 
and tho trikctonc, using a large exce&=i of the latter in high con- 
centration. The necessity of a high concentration to produce 
quantitative amounts of color in this reaction had been pointed 
out by Herzfeld,' and a series of experiments of our own had con- 
firmed this conclusion. We heated together in a boiling constant- 
level water bath a scries of 1 cc, of 0.1 per cent solution of amino- 
aci<l and O.o cc. of I jicr cent Kolution of triketohydrindone hydrate 



•Herzfolil: he cil. 



V. J. Harding and R. M. MacLean 223 

for periods of 5, 10, 15, and 20 minutes. The amount of color pro- 
duced in the test-tube heated for five minutes was taken as 
standard, and the results were plotted as before. Each amino- 
acid was considered separately (Figure 2). 

The effect of the higher concentration of the triketone was at 
once readily apparent. The amounts of color produced in the 
same time were much larger, and concordant experiments were 




Fig. 2. 



easily obtained except in the case of asparagine. It will also at 
once be noticed that the times of maximum development of color 
with the different amino-acids are much shorter than in the first 
series of experiments and lie closer together. Thus, in the first 
series, the times vary from twenty minutes (tyrosine) to seventy 
minutes (alanine), whereas in the second scries the variation is 
only from ten minutes (glycine) to twenty minutes (aspartic acid 
and tyrosine). The other amino-acids possess the time of maxi- 



2 24 Estimation of Amino-Acid a-Nitrogen 

mum development of color at fifteen minutes. The colors with 
aspartic acid and asparagine, however, were still very weak and 
still possessed a strong yellowish tinge which rendered their com- 
parison with the other acids extremely difficult. Other irregu- 
larities, too, were noticeable. The colors produced by aspartic 
and glutaminic acids f^ded very much more rapidly than those 
produced l)y the other amino-acids; that produced by aspartic 
acid being discharged completely in a dark room at the end of 
twelve hours. As these tw^o acids differed from the others only 
in the presence of a second carboxyl group, thus rendering them 
more acidic in character, the disappearance of the coloring matter 
on standing was put down to this cause. Ruhemann had pointed 
out that long contact with acids causes a decomposition of the 
coloring matter, giving a colorless solution, and experiments l)y 
us had proved the strong inhibitory effect of small amounts of 
organic acids on the production of the ammonium salt of diketo- 
hydrindylidene-diketohydrindamine from the interaction of ala- 
nine and the triketone. 

In order to test this assmuption a third scries of experiments 
was carried out in presence of a base which would neutralize the 
acidity of the second carboxyl group of aspartic and glutaminic 
acids. Such a base should not be strong enough to hydrolyze the 
triketohydrindene hydrat<?, as happens with the hydroxides of 
the alkali metals, and should not interfere with the reductions 
and condensations which take place in the reaction. The base 
chosen was pyridine. 

Inter aHion of amino-dcids and triketohydrindene hydrate in large 

excess, in presence of pyridine. 

One cc. of a 0.1 per cent solution of the amino-acid was added 
to 0.5 cc. of a 1 per cent solution of triketohydrindene hydrate*, 
and 0.2 cc. of pure, freshly distilled pyridine was added, and the 
mixture heated in a boiling water bath for varying intervals of 
time, as in the two previous series of experiments. The estima- 
tions of the relative amounts of coloring matter were carried out 
in a manner similar to the former experiments, each acid being 
compared with itself only. The curves in Figure 3 show the 
results obtained. They show quite clearly that the addition of 



V. J. Harding and R. M. MacLean 225 

the pyridine had achieved the desired result. They show that 
the reaction between amino-acids and triketohydrindene hydrate 
in presence of pyridine takes place rapidly, and reaches a maximum 
amount of coloration which remains constant for some minutes, 
except in the case of aspartic acid. This time of maximmn devel- 
opment of color is constant at about twenty minutes. Moreover, 
the addition of the pyridine had increased enormously the actual 
amounts of coloring matter produced, and the colors now did not 
fade appreciably in a short period of time. In the curves obtained 
with glycine and alanine 0.5 cc. of pyridine was used. 



IS 

1. 












? 

•^ 

K 

^ 










ifllanine. 




15 








^^- 




Jfspart/c Add 
tCifutQvthuc ftck 













^=r- '1, .-J 


lifrosin^ 


I 




H 




5 A 


9 a 


' a 


9 J 


• J» 



Time in f/linutes 
Fig. 3. 



A determination was now made of the relative amounts of 
coloring matter produced by four of the amino-acids, of whose 
purity we were quite certain, by dissolving equivalent amounts 
of each of them in water and using one as a standard. The fol- 
lowing table will illustrate the experiment and the result. Being 
in equivalent amounts, they should produce the same amount of 
coloring matter and thus give the same colorimeter reading. 

It will be seen that aspartic acid was still a little low in amount 
of coloring matter produced when compared with the other acids. 

As a result of further experiments, it was decided to reduce the 
amount of amino-acid nitrogen per cc. to 0.05 mgm., to reduce 



226 Estimation of Amino-Acid a-Nitrogen 



AMINO-ACID 



Alanine 

Glycine 

Aspartic acid. . . 
Glutaminic acid 



»ER CC. 


! PTRIOINE 

1 (PURK) 


TRIKETONE 
1.0 PER CEXT 


; COLORIMETER 
i READING 


mym. 


1 CC. 




CC. 




0.1 


0.2 




0.5 


1.00 


0.1 


0.2 




0.5 


0.99 


0.1 


, 0.2 




0.5 


1.17 


0.1 


0.2 




0.5 


1.01 






. 







the pyridine to 1.0 cc. of a 10 per cent solution in water, and to 
increase the triketohydrindene hydrate to 1.0 cc. of a 2 per cent 
solution. 

Method of estimation of amino-acid a-niirogen by u.'ie of triketo- 
hydrindene hydrate. 

One cc. of the solution to be estimated, containing not more than 
0.05 nigm. of amino-acid a-nitrogen and neutral to phenolphtha- 
lein, is mixed with 1 cc. of a 10 per cent aqueous solution of pure 
pyridine and 1 cc. of a freshly prepared 2 per cent solution of tri- 
ketohydrindene hydrate and heated in a rapidly boiling constant- 
level water bath for twenty minutes. At the end of that time the 
test-tube is removed, cooled, and diluted to a suitable volume, 
usually 100 cc, but if the amino-acid a-nitrogen is very small in 
amount, a correspondingly smaller dilution can be used. The solu- 
tion of coloring matter thus obtained is compared with the stand- 
ard color in the usual way, in a Duboscq colorimeter. 

Preparation of a standard color. 

The standard solution is prepared by dissolving 0.3178 of a 
gram of pure freshly crystallized alanine in a liter of distilled 
water. Such a solution contains 0.05 mgm. of nitrogen per cc. 
To prepare the standard color 1 cc. of the standard alanine solu- 
tion is heated for twenty minutes in a boiling water bath with 
0.5 cc. of a 10 per cent solution of pyridine and 1.0 cc. of a 1 per 
cent solution of triketohydrindene hydrate.^" At the end of that 
time the contents of the tube are cooled and diluted to 100 cc. 
This solution of coloring matter is used as a standard and will 
keep for twenty-four hours. The standard solution of alanine is 

^° The employment of a more concentrated solution of the ketone gives 
no further increase in the amount of coloring matter in the case of alanine. 



V. J. Harding and R. M. MacLean 227 

stable for three months. Attempts were made to prepare a per- 
manent standard color without success. A preparation of the 
pure coloring matter was made according to the directions of 
Ruhemann, and a dilute solution of it made in water. It was 
found, however, that when the dried coloring matter was used, 
it was not as freely soluble in water as the freshly prepared sub- 
stance, and that an aqueous solution of the pure coloring 
matter faded rapidly; it was much more stable in the presence of 
pyridine, but even then at the end of three months it had faded 
completely. There was thus no advantage to be gained in using 
a solution of the coloring matter prepared per se. The standard 
color prepared as directed is little or no trouble to prepare along 
with the determinations themselves. 



Determination of amin^'Ocid a-nitrogen in various amino-acids and 
comparison vnth the method of Van Slyke. 

A series of determinations was then made of the amino-acid 
nitrogen in various amino-acids, and the results were compared 
with theory in the case of the pure acid. A parallel series of 
determinations was made by the method of Van Slyke and the 
results were compared with the colorimetric determinations. In 
the case of the acids whose purity was doubtful, the method of 
Van Slyke served as a standard. The results are expressed in 
mgm. of nitrogen per cc. 



AMINO-ACID 



Glycine 

Alanine 

Valine 

Leucine 

a-Amido-n-c.aproic acid. 

Phenyl alanine 

i-Tyrosine 

Tiyptophane 

Histidine 

Aspartic acid 

Asparagine 

Glutaminic acid 



NITBOGEN PER CC. 


Theory 


Colorimeter 


Van Slyke 


fngtn. 


mgm. 


mgm. 


0.186 


0.189 


0.189 


0.100 


0.101 


0.102 


0.201 


0.201 


0.196 




0.104 


0.102 


0.160 


0.164 


0.161 


0.128 


0.126 


0.126 




0.084 


0.087 




0.090 


0.092 




0.085 


0.089 


0.105 


. 103 


0.104 


0.17:» 


0.178 




0.15S 


0.157 

1 


0.157 



2 28 Estimation of Amino- Acid a-Nitrogen 

The results show quite clearly the excellent agreement of the 
colorimetric method with theory and with the method of Van 
Slyke. It must be pointed out that the solutions of the amino- 
acids were made of just sufficient concentration to be estimated 
in the Van Slyke micro apparatus. To determine the amount 
colorimetrically, it was necessary to dilute those solutions from 
two to four times, so that our actual error of determination is less 
than appears in the table. 

The estimation of cystine, however, by this method gives 
results which are much too low. Moreover, the coloration devel- 
oped has an intense red color, which makes it almost impossible 
to compare it with the standard alanine. As a consequence of 
this it would be impossible to estimate the amino-acid a-nitrogen 
in a mixture of amino-acids containing a large proportion of 
cystine without a large error. Such cases, however, arc not very 
common, and the authors do not think that the amount of cystine, 
occurring in the hydrolysis mixture obtained from the majority 
of proteins, will introduce any serious error. To determine this 
point and to see if the method would estimate the a-nitrogen in a 
mixture of amino-acids in the proportions in which they occur in 
a native protein, an estimation of the amino-acid a-nitrogen in 
ereptone was made and compared with the Van Slyke method. 





- — - - 


NtPEROC. 


N« 


Colorimeter 


mqm. 

0.132 
0.139 


per cent 
8.5 


Van Slyke 


9.0 









It will be seen that the colorimetric method gives figures 
slightly lower than the Van Slyke method, but not low enough to 
invalidate seriously the result. The difference, however, is ca- 
pable of easy explanation. The colorimetric method gives too 
low a result in the presence of cystine. The Van Slyke method 
gives a result a little too high in the case of glycine and cystine. 
These three facts are sufficient, we think, to explain the difference 
of 0.5 per cent. 

A series of determinations of the amino-acid a-nitrogen in 
peptones from various sources was next carried out in order to 



V. J. Harding and R. M. MacLean 229 

test the efficacy of the method when dealing with partial hydrol- 
ysis products of proteins. The following table shows the results 
to be in good agreement with those obtained by the Van Slyke 
method. 



N3 PEB CC. 



ORIGIN OF PEPTOXB 



Colorimeter Van Slyke 



Meat (Merck) 0.062 , 0.061 

*'^ carnc'' (Schuchardt) I 0.035 0.031 

Precipitated by alcohol (Schu-; j 

chardt) 0.025 ■ 0.021 

Witte I 0.067 \ 0.067 



Determination of the maximal and minimal amounts of amino-acid 
a-nitrogen capable of estimation by this method. 

The early experimental results described in this paper (page 
226) had shown that although some of the amino-acids could be 
accurately estimated when the solution contained 0.1 mgm. of 
nitrogen per cc, yet other amino-acids could be estimated only 
when the nitrogen was only 0.05 mgm. per cc. This value is 
taken as the maximal value. To determine the minimal value 
two series of experiments were performed, one on an alanine 
solution, the other on an ereptone solution. A standard solution 
of alanine and a solution of ereptone were diluted by known 
amounts of water and colorimetric estimations made of the amino- 
acid a-nitrogen present in the diluted solutions. The coloring 
matter produced was diluted to a suitable volume so that the 
colorimetric readings did not differ from the standard by large 
amounts. The theoretical amounts given against the results 
obtained for ereptone solutions were obtained by calculating from 
the mean result of the colorimetric and Van Slyke determinations 
given on page 228. 

mQin, mQtn. tngm. tngtn. 

Alanine Found 0.050 0.025 0.012 0.005 

Perec Calculated 0.050 0.025 0.012 0.006 

Ereptone Found 0.065 0.032 0.016 0.008 

• Perec Calculated 0.067 0.033 0.016 0.008 



230 Estimation of Amino-Acid a-Nitrogen 

It will thus be seen that the method is accurate over a range 
of 0.05 mgm. to 0.005 mgm. per cc. Whether amounts less than 
0.005 mgm. can be estimated, we have not determined. At that 
concentration the amount of coloring matter produced is so small 
that it can only be measured with difficulty. Qualitatively, how- 
ever, it is possible to detect as little as 0.001 part of a mgm. of 
amino-acid a-nitrogen in 1 cc. of solution. This, in the case of 
alanine, means the detection of one part of the amino-acid in a 
little over 1,500,000 parts of water. 

The question naturally arises as to whether the reaction will 
proceed quantitatively in presence of other substances. From 
the results obtained with ereptone (page 228) and the various 
peptones (page 229), the method can be used for the determination 
of the amino-acid a-nitrogen set free in the hydrolysis of proteins. 
Its application to the analysis of urine and blood, however, is still 
attended with difficulties, ammonium salts and urea causing 
disturbances in the reaction. These points are being investigated, 
and the results will be reported shortly. 

SUMMARY. 

1. A method has been devised for the estimation of amino-acid 
a-nitrogen by the use of triketohydrindene hydrate and pyridine. 

2. The method possesses an accuracy equal to that of the Van 
Slyke method. 

3. It will estimate within the ranges of 0.005 mgm. to 0.05 mgm. 
amino-acid a-nitrogen per cc. 

4. The method is inaccurate, however, for cystine. 

5. The method is applicable to the determination of amino- 
acid a-nitrogen (in neutral solution) set free in protein hydrolysis. 



THE METABOLISM OF VEGETARIANS AS COMPARED 

WITH THE METABOLISM OF NON-VEGETARIANS 

OF LIKE WEIGHT AND HEIGHT. 

By FRANCIS G. BENEDICT and PAUL ROTH. 

{From ilic Nutrition Laboratory of the Carnegie Institution of Washington , 
Boston, and the Battle Creek Sanitarium, Battle Creek.) 

(Received for publication, January 15, 1915.) 

Perhaps no special dietetic regime is more successful in attract- 
ing adherents and holding them than is that of vegetarianism. 
While the modern conception of the word ''vegetarianism'^ is 
by no means as strictly applied to the exclusive use of vegetables 
as was formerly the case, and we have at the present time the 
lacto-vegetarians and the ovo-vegetarians, nevertheless there 
are a considerable number of individuals who regularly confine 
themselves to a strictly vegetarian diet. 

Large groups of people, particularly in foreign countries, are 
by long custom vegetarians. In this study, however, we are 
specially concerned with a considerable number of individuals 
who, notwithstanding the fact that they live among people 
partaking of an ordinary mixed diet, yet adhere closely to a 
vegetarian diet. This adherence to a vegetarian diet may be due 
to any one of several causes: First, to environment, heredity, 
training, or the habit of the household and parents; second, to a 
religious belief, in which the use of flesh for food is proscribed; 
third, as the result of some dietetic alteration incidental to the 
treatment of disease and a subsequent adherence to the vege- 
tarian diet; or finally, for physiologic or biologic reasons, in the 
belief that a non-flesh dietary is the natural or physiologic diet 
for human beings, as well as other primates. 

In the discussion of the supposed benefits of vegetarianism, 
considerable stress has been laid upon the remarkable endurance^ 

* I. Fisher: Yale Med. Jour., xiii, p. 205, 1907. See also loteyko and 
Kipiani: Rev. de la Soc. scient. d*hyg. alimentaire, ill, p. 114, 1906. 

231 



232 Metabolism of Vegetarians 

apparently shown by vegetarians over flesh eaters, the statements 
being made that the vegetarians live upon a distinctly lower 
metabolic plane, are not so highly stimulated as the flesh eaters, 
have a lower blood pressure, and, in general, that the metabolic 
activities are on a lower level. 

Fortunately, while it is wholly impossible to measure accurately 
and scientifically many of the indices of benefits or lack of benefits 
commonly cited by individuals, it is perfectly feasible to measure 
the metabolic plane by studying the respiratory exchange.^ 
Thanks to the kindness and interest of Dr. J. H. Kellogg of the 
Battle Creek Sanitarium, Battle Creek, Michigan, we were en- 
abled to make a series of observations on both -men and women 
vegetarian subjects, all of whom may be definitely classified as 
normal individuals. The greater number of the subjects were 
members of the staff of the Sanitarium. These individuals had 
been meat abstainers for varying lengths of time, the periods being 
stated in the list given below. It is safe to state that those 
designated in this list as strict vegetarians probably did not eat 
meat of any kind more than two or three times a year. 

Men. 

O. N. A. Strict vegetarian for 6 mos. 

B. N. C. Strict vegetarian for at least 10 yrs. 

V. E. H. Strict vegetarian for 1 yr. 

B. K. Practically a vegetarian for 2 yrs. Used meat probably 

once a month on the average. 
W. B. Vegetarian for 5 to 6 yrs. Has eaten meat not oftener 

than 2 or 3 times a yr. during that time. 
F. E. M. Strict vegetarian for IJ yrs. 

Dr. P. R. Strict vegetarian for many yrs., and practically all his life. 
E. H. T. Strict vegetarian for 8 mos. 
E. J. W. Strict vegetarian for over 20 yrs. 
L. IT. W. Except for a short period of 3 to 4 wks. (several yrs. ago) 

has been practically always a vegetarian. For several 

yrs. before the test had not eaten meat oftener than 

once a mo. 
T. H. Y. Strict vegetarian for over 5 yrs. 



* For a report of the most important study of the respiratory exchange 
of vegetarians thus far made, see W. Caspari: Arch, f. d. ges. Physiol., 
cix, p. 473, 1905. 



F. G. Benedict and P. Roth 233 

Women. 

Miss O. A. Strict vegetarian for 2 yrs. 
Mrs. E. B. Vegetarian diet for 3 mos. only. 
Miss J. U. B. Strict vegetarian for 2i yrs. 

Miss L. B. Very seldom ate meat at any time, practically none for 

10 yrs. ; none for 2 J yrs. before test. 
Dr. M. D. Strict vegetarian for 20 yrs., with the use of meat at very 

infrequent intervals, not more than 2 or 3 times a yr. 
Miss M. H. Strict vegetarian for 2 yrs. 
Miss M. J. Strict vegetarian for 6 wks. only. 

Miss L. K. Strict vegetarian for 1 yr. ; ate meat only on rare occasions. 
Mrs. A. L. Vegetarian for 5 or 6 yrs. Had eaten meat not oftener 

than 2 or 3 times a yr. during that time. 
Miss J. T. Strict vegetarian for 1 yr. 
Miss C. Z. Practically a life vegetarian. Had eaten no meat for 

14 yrs. 

The unit respiration apparatus was ufeed for this study, all due 
precautions being taken as to muscular repose, absence of food 
in the stomach, and a non-febrile temperature; every effort was 
made to approximate the simplest and most perfect physiological 
condition. Observations were carried out with some subjects 
on but one or two days, but frequently they were made on several 
days. All recorded values were derived from not less than two 
satisfactorily agreeing experimental periods. 

In comparing the results, it is obviously improper to obtain 
simply a grand average for all the experiments with vegetarians 
and compare this with an average for the experiments with non- 
vegetarians. .We have, therefore, adopted the plan of comparing 
the vegetarians with control individuals living on a mixed diet 
and of the same sex and approximately the same weight and height. 
The ages of our subjects ranged between 20 and 40 years, and we 
have only occasionally to deal with the possible influence of greater 
age; these cases will be specially noted. 

In the majority of instances it was not possible to select a nor- 
mal individual comparing exactly in height and body weight 
with a vegetarian; and, indeed, it would be wholly erroneous to 
select any one individual for this comparison. We have, there- 
fore, divided the comparisons into groups, using for each group 
several non- vegetarian individuals having approximately the 
same height and weight. Unfortunately the number of vegetarians 
studied, while considerably greater than ever l)ofore observed, 



2 34 



Metabolism of Vegetarians 



TABLE I. 



Comparison of the heat production of vegetarians and non-regctarians, 

{Experiments with men.) 



GROUP AND SrBJECT 



XUD K 
WEIGHT 



kgm. 



GUOUP I 

Vegetarian ' 

F. E. M 75.0 

Xo7i-vegetarian 

F. A. R I 74.4 ! 

Ghoip II 

Vegetarian 
K. H. T 64.7 i 

X on -vegetarian i 

F P R 65 1 

D.M 64.0 . 

C.Hori' III 

Vegvfarian 

L. H.W 60.0 

B. K ; 58.2 I 

Xon-regetariati 

\> . Ci. •! 60. o 

L. E. E 59.8 

Dr. S 5S.5 ; 

I). J. M 58.0 

II. F. T 57. S 

Guorp IV 

Vegetarian 

T. H. Y 50.2 

Xnjt-i'cgetarian 

A.L I 60.6 

II. B. U 60.5 

J. H.T 60.1 

W. F. M 60.1 

K. T. \V 57. S 

I'. F. J r)7.'J 

A. C. K 57.0 

Ciiori' V 

\'cgi tnrinn 

W. H. 1 59.3 

Xnn-rrgt In r inn 

H. IL A 02.3 

S. A. H 60. S 

H. B. H 60.5 

P. F.J 57.2 



IIKIGHT 
cm. 

164 
163 

170 



164 



HEAT PHODL'CTION WAX 24 HU**. 
(COMPUTKD) 



Total IVrkgm. P'J,,;','^?" 



cal. 



1698 



1704 



1499 



1451 



cal. 



22.7 



22.9 



23.2 



24.5 



cat. 



775 



782 



iO/ 



173 


1543 


23.7 


775 


171 


1651 


25.8 


838 


170 


1647 


25.9 


840 


179 


1530 


25.5 


810 


178 


1393 


23.9 


753 


175 


1746 


28.9 


919 


175 


1707 


28.5 


908 


181 


1331 


22.8 


716 


175 


1615 


27. S 


878 


179 


1348 


23.3 


733 


169 


1605 


> 

27.2 


861 


171 


1576 


26.0 


S29 


168 


1487 


21.6 


7s:^ 


171 


1748 


2<> . 1 


025 


168 


1632 . 


27 2 


n6:3 


169 


1172 


25.5 


S(HI 


167 


1616 


2S.3 


s.s:{ 


169 


ir>31 


26.9 


841 



770 



164 


1487 


23.9 


770 


165 


1460 


24.0 


7r>s 


168 


1487 


24.6 


7S3 


167 


1616 


2S.3 


SS3 



F. G. Benedict and P. Roth 



235 



TABLE I— Concluded. 



OHOrP AND bUBJErr 



N'UDE 
WEIGHT 



kgtn. 



IIKIGIIT 



cm. 



HSAT PRODrCTIOX PEB 21 HR8. 
(COMPUTED) 

Total Perkgni. ' ^^j^X"'* 



cal. 



cal. 



cal. 



Group VI 












Vcgetan'aJt 




1 








X^i. m k • L\ •••>«•>••.••••• 


55.2 


164 


1341 


24.3 


753 


Son-vcgcinrinii 












P F T 


57.2' 


167 


1616 


28. 3 


883 


«•&■ ''•■••■■••■••••••••a 


53.6 


160 


1455 


27.1 


831 


Group VII 












Vegetarian 












O.N. A 


00.4 


171 


1545 


27.9 i 


863 


Son^rcgitai'iun | 












T f) A 


57.1 


171 


1539 


27.0 


S41 


A. G. E 


57.0 
56.1 


169 
173 


1531 
1522 


26.9 
27.1 


841 




846 


C. II. H 


55 . 1 


169 


1421 


L'5.8 


79S 


Group VIII 












Yegeiur'xa^t 












B. N. C 


50.6 


179 


1510 


29.8 


893 


Son-vcgetan'an 












\ V G 


53.9 


175 


1453 


27.0 


826 


1j • I'jb a 


52.2 


174 


1541 


29.5 


896 


Group IX 












Vegetarian 












E. J. W 


50.0 


155 


1158 


23.2 


693 


Xon-vcgttarian 












I. A. r 


54.9 


150 


1612 


29.4 


906 


J. H 


46.3 


154 


1223 


26.4 


769 


Group X 












Vegetarian 












V r 11 


49.3 


163 


136.5 


27.7 


822 


\ari~rcgrtarian 












J . ti . G 


50 . 2 


164 


1425 


i:s.4 


848 


T M (' 


4S.5 


ia5 


1292 


26.0 


788 



i.s still too small to permit groupinpj, and, as a rule, only one 
vcgetiirian is used in each comparison. For this reason, we dis- 
tinctly decline to draw conclusions from the individual compari- 
sons, as they can be intelligently drawn only from the general 
picture. 

The comparison of the mal(» subjects is given in Table I. 
The subjects are here divided into ten groups, corresponding to 



236 Metabolism of Vegetarians 

eleven vegetarians, as, except in the case of Group III, no two can 
properly be classified in a single group. In choosing individuals 
living on a mixed diet for comparison, the sole bases for selection 
have been those of body weight and body height. In other words, 
every attempt has been made to secure a purely objective compari- 
son. An examination of this table shows that there is no strik- 
ing uniformity in the results, the metabolism of the vegetarians 
being at times somewhat lower than that of the control. Occa- 
sionally the vegetarians show a higher metabolic plane than any 
of the control subjects. On the other hand, the vegetarians Dr. 
P. R., Group VI, and E. J. W.,' Group IX, show a noticeably 
lower metabolism than either of the controls with whom thev 
are compared. 

Believing that a comparison of the average metabolism of each 
group of controls with the average metabolism of the vegetarian 
in the same group will give a better picture, the averages have 
been brought together in Table II. Here for the first time vfc 
are permitted to obtain an average of all the controls with an 
average of all the vegetarians, since we are now comparing vege- 
tarians with control individuals of approximately the same body 
weight and height. Accordingly in these final summary tables 
we may compare simply the heat per kgm. of body weight and 
per square meter of body surface. 

In the averages for the ten groups, the vegetarians show a 
higher metabolism than the controls in four cases. In the grand 
average the metabolism of the vegetarians (25.5 calories per kgm.) 
is approximately 4 per cent lower than that of the non-vegetarians 
(26.4 calories per kgm.). The same small difference is found 
when the comparison is made on the basis of heat per square 
meter of body surface. 

' The extraordinarily low values found with E. J. W. may in part be 
explained by the fact that he was the oldest man entering into this com- 
parison, being 58 yours old, and evidence has accumulated to show that age 
affects materially the intensity of the metabolism. Singularly enough, 
while the oldest woman entering into the comparison (Mrs. E.B.) is 53 years 
old, there is no noticeable difference in her metabolism over that of the 
other women vegetarians, although it is possibly of significance that she 
had been a vegetarian but 3 months. In both sets of comparisons the 
effort has boon made to select controls that would be least affected by 
differences in age. 



F. G. Benedict and P. Roth 



237 



From a cursory examinatiou of all the available data in this 
laboratory for normal individuals, it early became apparent 
that the metabolism of men may not properly be compared directly 
with that of women, and hence in this critical study of the metab- 
olism of vegetarians we have felt it important to compare men 
with men and women with women. The values found for the 
women vegetarians and their controls are given in nine groups 
in Table III. In only two groups, namely, Groups I and VIII, 

TABLE II. 

Summary of comparison of vegetarians and non-'Vegetarians, 

(Experiments with men.) 





OBOUP NO. 


1 

i 


HBAT PRODUCTION PBB 24 HBB. 
(COMPUTXD) 






1 Per kgm. of 
body weight 


Per square meter of 
body surface 







VegeUtriaDS 


Non- 
yegetariana 

eal. 


Vegetarians 
cal. 


Non- 
vegetarians 




1 cal 


cal. 


I. 




22.7 


22.9 


775 


782 


II. 




23.2 


25.1 


757 


818 


III., 




i 24.7 


26.3 


782 


831 


IV.. 




27.2 


26.8 


861 


846 


v.. 




24.6 


25.2 


776 


801 


VI.. 




1 24.3 


27.7 


753 


857 


VII.. 




27.9 


26.7 


863 


832 


VIII.. 


• ••••••••• 


29.8 


28.3 


893 


861 


IX.. 




1 23.2 


27.9 


693 


838 


X.. 


• ••••••••• 


27.7 

1 


27.5 

1 


822 


818 


Average 


25.5 


26.4 


798 


828 



is it possible to include more than one vegetarian. An examinatiou 
of the data shows that here again there is no striking uniformity 
in the direction of the values. Thus the two individuals, Dr. M. 
D. and Miss 0. A. in Group I, had a measurably higher metab- 
olism than the control, Miss H. H. Similarly, Miss C. Z. in 
Group II, had a considerably higher metabolism than any one of 
the three controls with whom she is compared. On the other 
hand, the vegetarians Miss J. U. B., Group V, and Miss L. B., 
Group VII, had a metabolism very considerably less than any one 
of the controls in the same group. 



TniJOUBNALfnr BXOLOOICALCHSSf10TBT, VOL. XX, NO. 3 



238 



Metabolism of Vegetarians 



TABLE III. 

Comparison of the heat production of vegetarians and non-veg^larians. 

{Experiments with women.) 



GHOUP A^a> SUBJECT 



Group I 
Vegetarian 

Dr. M. D. . . . 

Miss O. A . . . 
Non-vegetarian 

MissH. H... 
Group II 
Vegetarian 

MissC. Z.... 
Non-vegetarian 

MissS 

MissC. H... 

Miss A. K. . . 
Group III 
Vegetarian 

Mrs. E. B... 
Non-vegetarian 

Miss B. W . . 

Miss J. C 

Group IV 
Vegetarian 

Miss L. K. . . 
Nan -vegetarian 

MissM. W.. 

Miss M. P. . . 

Miss G. L. . . 

MissM. T... 

Miss F. K.. . 
Group V 
Vegetarian 

Miss J.U.B. 
Non-vegetarian 

Miss J. C 

MissF.E... 

Miss B 

MissR. M... 



KUDE 
WEIGHT 



kfftH. 



93.6 
90.2 

88.3 



67.2 

65.5 
63.4 
63.2 



58.0 

59.4 
55.1 



56.8 

58.6 
58.1 
55.0 
54.5 
54.1 



HEIGHT 



cm. 



165 
164 

161 



170 

171 
166 
171 



163 

162 
162 



166 

167 
168 
166 
164 
164 



HEAT PBODUCTIOK FEB 24 HBB. 
(COMPT7TED) 



Total 



eal. 



1765 
1756 

1591 



1521 

1426 
1413 
1402 



1415 

1546 
1363 



1365 

1429 
1518 
1480 
1359 
1262 



53.8 


160 


1215 


55.1 


162 


1363 


53.1 


162 


1391 


52.2 


158 


1415 


52.1 


162 


1353 



Per 
kgm. 

eoL 



18.9 
19.5 

18.1 



22.7 

21.8 
22.3 
22.2 



24.4 

26.0 
24.8 



24.1 



22.6 



Per Hquare 
meter 

eal. 



695 
708 

652 



747 

713 
722 
717 



769 

827 
764 



750 



24.4 


768 


26.2 I 


823 


27.0 


832 


25.0 


770 


23.4 


716 



694 



24.8 


764 


26.2 


799 


27.2 


823 


26.0 


787 



F. G. Benedict and P. Roth 



239 



TABLE UI— Concluded. 



OROUP AND SmUBCT , 

1 


NUDE 
WEIGHT 


1 

HBIOHT 

1 


HE.\T PRODUCTION PER 21 HR8. 
(COMPUTED ) 

Total ■ /-. ^-X" 




_ 

fcQtn. 


ctn. 


caL 


' cal. 


eal. 


(iuOUP VI 












Vegetarian 












MissM. H 


49.1 


151 , 


1178 


24.0 


712 


Non-vegetarian 




1 




1 




Miss R. A 


50.8 


155 


1293 


25.5 


765 


Miss G. F 


48.5 


155 


1233 


25.4 


754 


Group VII 












Vegetarian 












MissL. B 


47.0 


167 


1168 


24.9 


730 


Xon-vegetarian 












Miss I. B 


50.1 


166 


1235 


24.7 


737 


AvLiDo JLJ m Xs*********** 


46.7 


164 


1336 


28.6 


838 


Group VIII 












Vegetarian 












A * X X • ^\ m JLJ •■••••■••••• 


44.9 


159 


1272 


28.3 


815 


Miss M. J 


44.8 


157 


1189 


26.5 


767 


X on-vegetarian 












Miss H. T 


45.0 


159 


1393 


30.9 


896 


Miss J 


43.0 


159 


1158 


26.9 


766 


Group IX 




1 








Vegetarian 












Miss J. T 


40.0 


, 168 


1269 


31.7 


881 


\on -vegetarian 




1 








Miss A. C 


42.6 


1 165 


1168 


27.4 


779 



As noted with men, a comparison of the average values for 
women, irrespective of weight and height, may not properly be 
made; nevertheless the values for the separate groups may be 
averaged and thus a grand average obtained which gives a reason- 
ably correct picture of the probable relationship between the 
vegetarian and non-vegetarian women. These averages, which 
are calculated on the basis of per kgm. of body weight and per 
square meter of body surface, are given in Table IV. The 
vegetarians show a higher metabolism than the controls in four 
of the nine groups. In the grand average the heat production 
of the vegetarians on the bases of per kgm. of body weight and 
lH*r square meter of body surface is but slightly less than the 



240 



Metabolism of Vegetarians 



heat production of the non-vegetarians, the difference between 
the two classes being a little less than 2 per cent. 

From these data we may conclude that the male vegetarians 
have a slightly less metabolism per kgm. of body weight and per 
square meter of body surface than have the individuals living 
on a mixed diet with whom they are compared. This difference 
is so small, however, that as a general picture no essentially 
striking difference is apparent between male vegetarians and non- 
vegetarians. Certainly there is nothing to warrant the belief 



TABLE IV. 



Summary of comparison of vegetarians and non^vegetarians. 

(Experiments vnth women,) 



UROCP NO. 



BXAT PBODUCnON PBB 24 HB8. 
(COMPUTBO) 



Per kgm. of body weight 



VegeUruuui 



eal. 

1 19.2 

II 22.7 

III 24.4 

IV 24.1 

V 22.6 

VI 24.0 

VII 24.9 

VIII 27.4 

IX ; 31.7 

I 
I 

Average 24.6 



Per square meter of body 
surface 



Non- < 
letarians 


1 VegetariaoB 
: col. 


aU. 


18.1 


702 


22.1 


! 747 


25.^ 


769 


25.2 


750 


26.1 


694 


25.5 


71? 


26.7 


730 


28.9 


791 


27.4 


881 


25.0 


753 

1 



Non- 
yegetariane 

cal. 

652 
717 
796 
782 
793 
760 
788 
831 
779 

766 



that the male vegetarian subsists upon a materially lower met- 
abolic plane. With the female vegetarians, the slight difference 
in metabolism shown by the male vegetarians entirely disappears. 
We may, therefore, fairly conclude that living upon a vegetarian 
diet for a longer or shorter period docs not fundamentally alter 
the basal gaseous metaboUsm. 

One of the particular reasons for instituting a study of the me- 
tabolism of vegetarians is to note if the body of the vegetarians 
contains a larger proportion of readily combustible carbohydrate 



F. G. Benedict and P. Roth 241 

(glycogen) as the result of their special diet. Obviously vegetari- 
ans, in order to secure the total energy required for the day, 
must secure a larger proportion of energy from carbohydrate 
than do individuals living on a mixed diet; for it is much easier 
to obtain a larger proportion of animal fat than of vegetable fat. 
It was thought possible that vegetarians, subsisting for a long 
time on a diet rich in carbohydrates, would have acquired a stor- 
age of glycogen above that of an individual living' on a mixed 
diet, and that consequently in a quiet, resting condition, twelve 
hours after the last meal, they would show a katabolism with a 
larger proportion of carbohydrate. This would obviously be 
indicated by the respiratory quotient. 

In any comparison of the respiratory quotients it should be 
taken into consideration that tjiis factor is extremely difficult 
to determine with great accuracy, since all the errors incidental 
to the determination of both the carbon dioxide production and 
the oxygen consumption affect the values for the respiratory 
quotient. One must, therefore, speak somewhat guardedly in 
comparing respiratory quotients. On the other hand, we have 
available respiratory quotients obtained upon twenty-two 
vegetarians (and here it is unnecessary, for the present at least, 
to make distinctions between male and female vegetarians and 
between individuals of different height, weight, and age). We 
may properly compare these values with respiratory quotients 
previously determined upon 132 individuals subsisting on a mixed 
diet. In such a comparison deductions may reasonably be drawn 
from an average quotient. The average respiratory quotient 
found with the 22 vegetarians (i.e., 11 men and 11 women), 
was 0.83, while the average quotient found with 132 individuals 
subsisting on a mixed diet (77 men and 55 women) was 0.81. 
This diflference is slight and is wholly incomfpatible with the 
l)elief that vegetarians, when in the post-absorptive condition, 
have available any considerably larger proportion of easily com- 
bustible carbohydrate material than have non-vegetarians. 



THE METABOLISM OF ATHLETES AS COMPARED WITH 

NORMAL INDIYIDUALS OF SIMILAR 

HEIGHT AND WEIGHT. 

By FRANCIS G. BENEDICT and H. MONMOUTH SMITH. 

(From the Nutrition Laboratory of the Carnegie Institution of Washington, 
Boston, and the 'Chemical Laboratory of Syracuse University, 

Syracuse.) 

(Received for publication, January 15, 1915.) 

The special influence of athletic training upon the basal rest- 
ing metaboUsni; particularly of the fully trained athletic con- 
dition, has received scant attention in studies of normal metab- 
olism. It is true that in the researches carried out by the Zuntz 
school, especially those in which the work of marching was 
studied, the influence of training upon the muscular efficiency 
for external work was considered; furthermore, certain writers 
have recognized the desirability of indicating the general mus- 
cular development of their subjects, and we find occasional com- 
ments in the literature upon the relationship between muscular 
condition and the basal metabolism. 

Thus Speck^ states that muscular individuals consume more 
oxygen and produce more carbon dioxide than do non-athletic 
persons under similar conditions. In the various sunmiaries 
of measurements of the basal metabolism, Magnus-Levy and 
Falk,* and Loewy* have noted the muscular condition of their 
subjects. Magnus-Levy and Falk* conclude that three individuals 
with special muscular development showed a basal metabolism 
which was no greater and possibly somewhat smaller than other 

* C. Speck: Physiologie des menschlichen Athmens, Leipsic, 1892, p. 224. 
' A. Magnus-Levy and E. Falk: Arch.f. Anat. u. Physiol., Supplement, 

p. 321, 1899. 

* A. Loewy: Oppenheimers Handbuch der Biochemie, Jena, iv, p. 179, 
1911. 

* Magnus-Levy and Falk: loc. cit., p. 363. 

243 



244 Metabolism of Athletes 

individuals; per unit of body weight they showed a somewhat 
smaller metabolism than the average individual. 

In a study made of the metabolism of the vegetarian athlete, 
Karl Mann, Caspari' concluded from the higher basal metabolism 
tbat this individual had a much more powerful musculature 
than another subject of essentially the same weight. Mann 
had a height of 163.5 cm. and a body weight of 61.175 kgm. 
Herr B., with whom he was compared, was 175 cm. in height 
and weighed 63.18 kgm. 

Benedict and Carpenter,® in reporting a number of experiments 
on the mctaboUsm of normal individuals, included incidentally 
certain comparisons of four athletes with a number of non- 
athletes having approximately the same body weight and height. 
No special effort was made to study the metabolism of athletes in 
particular, save in the observations with a professional bicycle 
rider. The other three subjects were college athletes, one of 
whom was a trained bicycle rider, another a football player, and 
the third especially proficient in gynmasium work. All four 
athletes showed a distinctly lower metabolism, both per kgm. of 
body weight and per square meter of body surface, than did the 
normal individuals with whom they were compared. As was 
pointed out by the writers, these experiments were not planned 
primarily with the idea of comparing the athletes with a group 
of non-athletes. Neither were the experiments made under 
ideal conditions for comparison; for it has been clearly shown that 
the best method of determining the basal metabolism of different 
individuals is with the subject in the post-absorptive state and 
with complete muscular rest. These individuals were compared 
on an entirely different basis, as they were not invariably in the 
post-absorptive state and were all sitting more or less quietly 
in the calorimeter chamber. Nevertheless the fact that the 
results were so uniformly low with the four subjects cannot be 
overlooked, and, as the authors themselves stated, a special 
research on this question was needed before an intelligent com- 
parison of the basal metabolism of athletes and non-athletes 
could be made. 

» W. Caspari: Arch. /. d. ges. Physiol. ^ cix, p. 473, 1905. 
• F. G. Benedict and T. M. Carpenter: Carnegie Institution of Wash- 
xngton Publications, No. 126, p. 236, 1910. 



F. G. Benedict and H. M. Smith 245 

As the experimental data for normal values accumulated in 
the Nutrition Laboratory, it became evident that the body 
composition probably had a considerable influence upon the 
metabolism. When all the results were superficially inspected, 
the general picture appeared to be that distinctly fat individuals 
showed a low metabolism per kgm. of body weight and per square 
meter of body surface. It seemed therefore of double interest 
to study the effect upon the basal metaboUsm of a greater than 
normal proportion of protoplasmic material such as was pres- 
ent in the body of the trained athlete. 

No opportunity presented itself for the study of a group of 
trained athletes until the winter of 1912-1913, when one of us 
made a series of observations in the chemical laboratory of 
Sjrracuse University. The subjects of this study were selected 
from among the students of the University; during the experi- 
ments they continued to eat at their respective college clubs, 
confining themselves entirely to the plainer articles of diet. Prac- 
tically none of them were using tobacco, although it is possible 
that some occasionally indulged. 

The sports engaged in were crew practice, 16 pound hammer 
throw and shot put, one-quarter, one-half, and one mile runs, 
100 yard dash, relay team, and basket ball; during the previous 
year several of the men had played football. While they were 
takmg part in their respective sports daily, they were not in that 
degree of training which might be expected of a professional prize 
fighter or a 6 day bicycle rider. Neither were they all of an equal 
degree of proficiency; a few were in their first season of regular 
sports, but others had been competing not only during their 
entire college course but had competed in interscholastic meets 
before entering college. One of the men a few weeks later was 
the first winner in an event at the Olympiad in Stockholm, an- 
other in an intercollegiate meet at Philadelphia, and others won 
in their events in the intcrclass meets. Broadly speaking, they 
were all active young men, participating regularly from one to 
three hours a day in sports. 

The results obtained with these athletes are compared in Tabic 
I with observations made upon non-athletes of similar height and 
weight. With the values obtained with the college athletes 
other values are included which were secured with the professional 



246 



Metabolism of Athletes 



TABLE I. 
Comparison of the heat production of athletes and normal non-<tthletic men, 



OHOCP AXD SUBJECT 


NUDE 
WBIQHT 


HEIGHT 


HEAT PROOUCnON PER 24 HRS 
(COMPUTED) 

ToUl Perkgm. ^^j^-^" 


Group I 


KQttlm 


cm. 


cal. 


cal. 


cal. 


AthleU 












W. S 


88.5 


165 ' 


2017 


22.8 


823 


Non-athlete 












V^ • ^ • X V *••«•••••••••• • 


85.8 


171 


1827 


21.3 


761 


Group II 












AthleUs 












V • AA 9 XVtt •••••••••••••• 


82.2 


187 


1978 


24.1 


849 


D. H. W 


82.1 


186 


2034 


24.8 


873 


M. H. K 


79.0 
78.9 


188 
184 


1944 
2126 


24.6 
27.0 


856 


E. G 


940 


Non-athletes 




A • ^* • SJ • •«••••••••••• 


83.1 


183 


1802 


21.7 


770 


W. A. M 


78.0 


183 


1816 


23.3 


807 


Group III 












Athlete 












S. • \J« Xv« .•....•..•••• 


74.0 


179 


1914 


25.9 


882 


Non-athletes 












WIT 


74.2 


183 


1770 


23.9 


816 


^^ ■ XJ • ft^« •••■■••••••••• 


71.1 


179 


1700 


23.9 


806 


Group IV 












Athletes 












V^' • X.^ • X\ ■«■•■•••■••■•. 


74.0 


173 


1908 


25.8 


879 


H. R. W 


73.9 


175 


1842 


24.9 


848 


Non-athletes 












Dr. M 


75.9 


175 


1877 


24.7 


849 


T P P 


73.7 


169 


1526 


20.7 


706 


H. W. E 


73.0 


168 


1559 


21.4 


725 


Group V 












Athlete 












PDF 


71.2 


176 


1810 


25.4 


858 


ATon-ofA/cfcs 












C. B. S 


71.1 


179 


1700 


23.9 


806 


T TJ TJ 

V • XX • XX •••••••••••••• 


69.1 


171 

1 


1634 


23.6 


789 


B. A. W 


67.9 


174 


1945 


28.6 


949 


Group VI 












Athlete 












M. A. M 


66.0 


176 


1695 


25.7 


843 


flv V ^B ■ ^ ^V ■ ^B * ^B V •■■■■■#•9999 





F. G. Benedict and H. M. Smith 



247 



TABLE I— Concluded. 



CROUP AND SUBJECT 



Non-athletes 
B. A. W.... 
F. P. R.... 

V • w • v^« .... 

R. G 

Group VII 
Athletes 

M. Y. B. . 

R. D. S. . . . 
Non-athletes 

F. P.R.... 

J. J. C 

E. H.T 

D. M 



R.G 

xl. v/. 15 

Group VIII 
Athlete 

W. F. M 

Non-athletes 

K. H. A 

J. R 

Dr. S 

fiROUP IX 

Athlete 

C. J. D 

Son-athletes 

Dr. P. R 

I A F 

M. B 

Group X 
AthleU 

W • At O. 

Non-athletes 
E. T. W... 
P. F. J... 
L. D. A.. 
A. G. E.. 
C H. H.. 



i w'SSSt i ^'O" 



A vJL • %M • 1^ •••••••••••••■ » 



• • • • • 



HEAT PBODUCTION PXB M BBS. 
(COMPUTBD) 



hgm. 



56.7 



cm. 



160 



Total |P.rk«n. iP«;;,X" 



cat. 



eal. 



1524 



26.9 



eal. 



67.9 


174 


1945 


28.6 


949 


65.1 


173 


1543 


23.7 i 


775 


65.0 


175 


1585 


24.4 ' 


796 


62.7 

1 


173 


1590 


25.4 


820 


63.6 


172 


1677 


26.4 


856 


, 63.5 


170 


1619 


25.5 


826 


65.1 


173 


1543 


23.7 ' 


775 


65.0 


175 


1585 


24.4 ' 


796 


64.7 


170 


1499 


23.2 


757 


64.0 


171 


1651 


25.8 


838 


; 63.7 


170 


1647 


25.9 


840 


1 62.7 


173 


1590 


25.4 


820 


62.0 


173 


1653 


26.7 


856 


62.4 


180 


1816 


1 
29.1 ; 


936 


66.4 


182 


1654 


24.9 


819 


66.0 


182 


1679 


25.4 


835 


58.5 


181 


1331 


22.8 i 


716 



838 



55.2 


164 


1341 


24.3 


753 


54.9 


156 


1612 


29.4 


906 


53.6 


160 


1455 


27.1 


831 


56.3 


169 


1562 


27.7 ' 


863 


57.8 


169 


1472 


25.6 


800 


57.2 


167 


1616 


28.3 


883 


57.1 


171 


1539 


27.0 


844 


57.0 


169 


1531 


26.9 


841 


55.1 


169 


1421 


25.8 ' 


798 



248 Metabolism of Athletes 

bicycle rider, M. A. M., who was employed as a subject by Dr. E. 
P. Cathcart^ at the Nutrition Laboratory in the winter of 1911- 
1912. 

In making such a comparison several difficulties are immedi- 
ately encountered. In the first place the athletes were not all 
of the same height and weight; hence it is necessary to compare 
them with the non-athletes individually and not collectively. 
Furthermore, when an individual of a given weight and height 
is to be compared with other individuals of the same weight and 
height, it is obvious that unless a very large number of control 
individuals are available to select from it is extremely difficult 
to obtain adequate normal data. A certain amount of difference, 
both in weight and height, must therefore be allowed in the col- 
lection of a suitable number of individuals for comparison. We 
have attempted to select the values impartially, making our choice 
of controls on the basis of weight and height as nearly as possible 
identical with those of the athletes. The values for all these 
comparisons are taken from the large table given in the paper 
on the basal gaseous metabolism of men and women which was 
recently pubUshed by us in conjunction with Emmes and Roth.* 
Further details regarding these subjects may be obtained by re- 
ferring to the original pubUcation.' 

In the comparison of the metabolism of athletes and non-ath- 
letes the value? are divided into ten groups (see Table I), the 
arrangement being in the order of the decreasing body weight. 
The initials, body weight, and height of each subject are given. 
Since individuals of approximately the same height and weight 
are compared, the total heat production, calculated for twenty- 
four hours, is recorded. It is thus seen that the comparison be- 
tween the two classes in each group may properly be made upon 
the basis of total heat production alone. On the other hand, since 
it is the habit of many writers to consider in such comparisons 
the heat production per kgm. of body weight per twenty-four 

^ F. (;. Benedict and E. P. Cathcart: ibid.y No. 187, 1913. 

• F. G. Benedict, L. E. Emmes, P. Roth, and H. M. Smith: this Journal, 
xviii, p. 139, 1914. 

* Two corrections, which slightly affect the values in the original table, 
are a decrease of 5 cm. in the height of Dr. M. and an increase of 3 kgm. 
in the weight of Dr. R. 



F. G. Benedict and H. M. Smith 249 

hours, and particularly the heat production per square meter of 
body surface as computed by the Meeh formula per twenty- 
four hours, these are likewise included in the table. 

An examination of Table I shows that the first eight athletes, 
1.6., those in Groups I, II, III, and IV, had with but one exception 
a greater total heat production per twenty-four hours on the three 
bases of comparison than any of the non-athletic individuals 
with whom they are compared. These differences are for the 
most part of considerable magnitude, I>eing from 200 to 300 
calories. The professional athlete, M. A. M., in Group VI shows 
a considerably higher heat production per twenty-four hours 
than do three of the four non-athletes with whom he is compared. 
The fourth uon-athlete, B. A. W., had a very high metabolism of 
1945 calories per twenty-four hours, 28.6 calories per kgm. of 
body weight per twenty-four hours, and 949 calorics per square 
meter of body surface per twenty-four hours. This subject, 
although classed as a non-athlete, was an assistant machinist in 
the Nutrition Laboratory, and accustomed to doing comparatively 
heavy machine work, piping, and boiler work. Although a well de- 
veloped and somewhat muscular individual, he could not possibly 
be classified as a trained, hardened athlete. He is retained in 
this comparison to show the possibilities of an increased metab- 
olism in individuals without a distinctly athletic training. It 
should be stated, however, that this is by far the most extra- 
ordinary case we have ever noted of so great an intensity of 
metabolism in an individual of this height and weight. 

The general picture presented in Group VII, in which M. Y. B. 
and R. D. S. are compared with seven non-athletes, shows no 
diflference in favor of either class. On the other hand, with the 
athletic subject W. F. M. in Group VIII there was a very large 
increase in the metabolism over that of the three non-athletes in 
the same group; in fact, this athlete had the highest metabolism 
per kgm. of body weight per twenty-four hours of any athlete 
studied. 

In Group X, in which the one athlete had the smallest body 
weight of any of the athletes studied, th(»re was practically no 
difference in the metabolism of the two classes. 

It is clear from this comparison, therefore, that the heaviest 
athletes, with a lx)dy weight of 65 kgm. or ov(»r, had almost 



250 Metabolism of Athletes 

invariably a pronounced increase in the metabolism above that 
of the non-athletic individuals of approximately the same body 
weight and height with whom they are compared. On the other 
hand, with athletes below 65 kgm., with the single exception of 
W. F. M., the difference is by no means so striking. It is evident 
that further observations with athletes of this weight are much 
needed to complete the study. Nevertheless, the evidence Is 
sufficiently plain to show that the observations of Benedict and 
Carpenter were distinctly deficient, since their four subjects all 
showed a considerably lower metabolism per kgm. of body weight 
and per square meter of body surface than did the normal individ- 
uals. In this study, however, we find that none of the athletes 
show a characteristically lower metabolism than the non-athletes, 
and that a very large proportion of them show a considerable 
increase in the metabolism. 

While the method of presentation outlined has been used, as 
no other method is available, we expressly wish to deny the legality 
of basing conclusions upon any individual comparisons or, indeed, 
the comparisons in any given groups. In Table II, therefore, we 
give the average value for each group and also a grand average 
for all of the athletes and non-athletes. Owing to the unequal 
number of athletes and non-athletes, it is somewhat inconsistent 
to find the grand averages and draw deductions from these values. 
On the other hand, an inspection of the general picture as shown 
in Table II is perfectly legitimate and this must be the only de- 
termining factor in drawing final conclusions. 

Of the ten groups which are averaged in Table II, we find that 
the heat per kgm. of body weight is greatest with athletes in all 
but two instances, i.e., in Groups V and IX, in which there was no 
difference. In none of the comparisons did a non-athletic group 
have a higher metabolism than the athletes with whom they were 
compared. The final average also shows a higher metabolism 
for the athletes of 26.0 calories per kgm. of body weight per twenty- 
four hours as compared with 24.4 calories per kgm. for the non- 
athletes. On the basis of per square meter of body surface we 
again find that the athletes exceed the metabolism of the non- 
athletes in all the groups, although the increases in Groups V, VI, 
and IX are insignificant. The general average on this basis 
shows a metabolism of 863 calories per square meter of body sur- 



F. G. Benedict and H, M. Smith 



251 



face per twenty-four hours for the athletes as against 807 calories 
for the non-athletes. As has already been stated, the general 
averages must be taken with some reserve since an unequal num- 
ber of subjects are compared in the two classes. From the general 
picture of the comparison, however, the conclusion may be fairly 
drawn that athletes have a somewhat higher metabolism, both 
per kgm. of body weight and per square meter of body surface, 
than do the non-athletes with whom we have compared them. 

TABLE II. 

Summary of comparison of athletes and non-athletes. 







HEAT PBODUCnOX PER 24 HR8. 
(COMPUTED) 




GROUP NO. 


Per ksm. of body weight 


Per square meter of 
body surface 




Athletes 
eal. 


Non-athletes 


1 
Athletes 

eal. 


Non-ethlctes 




eal. 


eal. 


I 

II 

Ill 

IV 

V 

VI 


22. S 
25.1 
25.9 
25.4 
25.4 
25.7 
26.0 
29.1 
26.9 
27.7 

26.0 


21.3 
22.5 
23.9 
22.3 
25.4 
25.5 
25.0 
24.4 
26.9 
26.7 

24.4 


823 

880 

882 

864 

858 
! 843 ' 
, 841 

936 
i 838 
': 863 

863 


761 
789 
811 
760 
848 
835 


VII 


812 


VIII 


790 


IX 


830 


X 


833 


Average 


807 



The possible objection may be raised to these observations . 
that although the athletes were studied in the morning, twelve 
hours after the last meal, yet they had been engaged in active 
nuiscular exercise the day before, and there may have been a 
prolonged after-eflfect of work. Benedict and Cathcart*® have 
indeed shown a considerable after-eflfect of work, lasting for several 
hours after the completion of severe bicycle riding, but it is hardly 
to l>e supposed that the after-effects of work carried out the day 
iMjfore would so persist for over twelve hours as to account for 
the. larger metabolism noted with the majority of these athletes. 
We believe we are therefore justified in concluding that the greatly 

1" Benedict and Cathcart: loc, cit,, p. 163. 



252 



Metabolism of Athletes 



increased proportion of active protoplasmic tissue present in the 
trained, hardened athlete is alone sufficient to account for the 
increase in the metabolism, and that this is not only an ab^lute 
increase, but from the nature of the comparison the metabolism 
is likewise increased per kgm. of body weight and per square 
meter of body surface. It would thus appear that the increase 
in the metabolism noted with athletes points strongly towards 
the earlier conception that the katabolism of the body is pro- 
portional not to the siuface of the body, but to the active mass of 
protoplasmic tissue. 



A COMPARISON OF THE BASAL METABOLISM OF 

NORMAL MEN AND WOMEN. 

Bt F. G. benedict and L. E. KMMES. 

{From the SiUrition Laboratory of the Carnegie In^itution of Washington, 

Boston.) 

(ReceiTed for publication, January 15, 1915.) 

As early as 1843 we find observations made by Scharling' 
in which both sexes were studied as to their metabolism and 
respiratory exchange. The values found by Scharling and 
recalculated by Sond^n and Tigerstedt- on the basis of carbon 
dioxide per hour in grams show that with a girl nineteen years 
of age there was a considerably less carbon dioxide excretion 
per hour than with either of two men or a sixteen year old boy. 
Scharling concludes from this single observation that there is 
a greater production of carbon dioxide with men than with women 
of the sanie age. 

The investigation of AndraJ and Gavarret' was very much more 
extended, inasmuch as thirty-seven men and twenty-two women 
were studied. Andral and Gavarret conclude that throughout 
the whole of life there is a greater production of carbon dioxide 
by men than by women, but that there is a proportionally larger 
production between the ages of sixteen and forty, when men, 
as a rule, produce about twice as much carbon dioxide as do 
women. Unfortunately Andral and Gavarret do not give the 
body weight of their subjects, and we are thus unable to compare 
them on the popular bases of per kgm. of body weight juid per 
square meter of body surface. 

In their classical monograph on the respiratory exchange and 
total metabolism of men, Sonddn and Tigerstedt^ published an 

* E. a. Scharling: Ann. d. Chem. u. Pharm.^ xlv, p. 214, 1843. 

* K. Sond6n and R. Tigerstedt: Skand. Arch. f. Physiol., vi, p. 54, 1895. 

* G. Andral and J. Gavarret: Ann. dc Chim. ci dc Phys., Series 3, viii, 
p. 129, 1843. 

* Sond^n and Tigerstedt: loc. cit., p. 58. 

253 

THE JOURNAZ. OF BIOLOOXCAL CH£UI8TRT, VOL. XX, NO. 3 






" y 



254 Metabolism of Normal Men and Women 

extensive scries of observations on both men and women in which 
the large respiration chamber in Stockholm was used. These 
results are comparable, although the observations were made 
under such conditions as to exclude them for use as indices of 
basal metabolism. Computing the values both for body weight 
and body surface, they conclude that in youth the carbon dioxide 
production of boys is considerably greater than that of girls of 
al)out the same age and the same body weight, but with increas- 
ing age this difference gradually becomes less and less and finally 
in old age it disappears entirely. It must be noted here that the 
authors specifically state that it appears to them that new 
experiments are necessary before this problem can be completely 
solved. 

In 1899 Magnus-Levy and Falk'' pubUshed an extended series 
of observations on both men and women in which the Zuntz- 
Geppert respiration apparatus was e^iployed. Although Johans- 
son^ had shortly before emphasized the importance of controlling 
muscular repose and had outlined his experience in the voluntary 
exclusion of muscular activity, these observations of Magnus- 
I-»evy and Falk represent the first comparative observations made 
upon both men and women in which particular attention was 
given to complete muscular rest; hence they are more perfectly 
comparable with our experiments than any series published pre- 
vious to 1899. The series with males comprise observations on 
sixteen boys, ten men between twenty-two and fifty-six years of 
age, and five men sixty-four years old and over. The series 
with women include observations on nine girls, fifteen women 
l)etween seventeen and fifty-seven, and seven women seventy-one 
years or older. The data as to the age, weight, and height 
are most carefully recorded and general comments made regard- 
ing the body condition, respiration rate, and pulse rate. The 
authors have likewise computed the values per kgm. per minute 
and per square meter of body surface per minute. In their com- 
parisons of the values obtained with men and women on the 
basis of per kgm. of body weight, they conclude that in middle 
life the gaseous metabolism of women is approximately the same 

* A. Magnus-Levy and E. Falk: Arch.f, Anat, u. Physiol, y Physiol. Abt., 
Supplement, p. 314, 1899. 

• J. 10. Johansson: Skand. Arch.f. Physiol. ^ viii, p. 85, 1898. 



F. G. Benedict and L. E. Emmes 255 

as that of men of the same age and body weight. With children 
and old men and women, the females have a slightly less (5 per 
cent) metabolism than the males. The authors also point x)ut 
that, owing to the larger proportion of body fat, females would 
have a metabolism per unit of active protoplasmic tissue greater 
than would men. 

Although since the experiments of Magnus-Levy and Falk, a 
large number of observations have been made on the metabolism 
of men, there have been relatively few observations on normaj 
women. These include those reported by Johansson,^ which 
were made with unusual care as to absolute muscular repose, and 
a few desultory studies by members of the Zuntz school. The 
experiments of Benedict and Carpenter* with two women in the 
respiration apparatus in Middletown, Connecticut, were not 
made with complete muscular rest; a considerable nmnber of 
observations upon non-athletic men were secured, however, under 
similar conditions of muscular activity, and the results were 
therefore comparable with those obtained with the women sub- 
jects. Both women showed a remarkably lower metabolism 
]>er kgm. of body weight and per square meter of body surface 
than did the men with whom they were compared. 

It is thus seen that in the last sixteen years no extensive com- 
l)arative study of the metabolism of normal men and women has 
appeared, and the last two great studies, namely, those of Sond6n 
and Tigerstedt and of Magnus-Levy and Falk, are distinctly 
at variance with each other as to the final conclusions. 

In accumulating normal material for comparison with patho- 
logical data in this laboratory and elsewhere, results obtained with 
eighty-nine men and sixty-eight women have been gathered 
together and presented in two tables in a paper recently published.-' 
It is our purpose here to discuss the values included in these 
tables and, as far as possible, to compare the metabolism of men 
and women. 



^ Johansson: ibid.f xxi, p. 1, 1909. 

•F. G. Benedict and T. M. Carpenter: Carnegie Institution of Wash- 
ington Publications, No. 126, p. 238, 1910. 

• F. G. Benedict, L. E. Eninies, P. Roth, and H. M. Smith: this Journal, 
xviii, p. 139, 1914. 



256 Metabolism of Normal Men and Women 

From the foregoing paper by Benedict and Smith,'" it is obvious 
that the normal men in this table who are distinctly classified as 
athletes cannot properly be used for comparison with women, 
and hence a grand average of the results for men cannot properly 
be compared with an average of the results for women. Since, 
however, such a comparison is not without interest, wc give in 
Table I an avera^ of all of the values found with the eighty-nine 
men and the sixty-eight women, Perhaps the most striking 
feature of this comparison is the fact that although the total 
metabolism, as measured by the carbon dioxide production and 
oxygen absorption, is greater with men than with women, the 
body weight is also considerably greater and likewise the height. 







TABLE r 










Average valuta for me 


abolUm of normal men 


and 








i "i ^ 


~| 


i 


S ' E 




"," 


ER^tHIB. 

CUTED) 




S 


e 


h i 






Jh 




»oJ ' & 


S 


^ 


03 <• 








; i ; 1 sg 


g^ 




°« e 






^ £S 






IP 


s 

E 


Pv 


3 


1 


if iff 
i' r- 




HT,.-. km- 


tm. «.;«. 


«. , M. 




col. 


nl. cal. 


men 


26 8 1 2S M.3 
27. 2 5 54.5 


172'l96235 


3.04:3.66 


ei 


1638 


25.5 S32 


women. . - 


163158 


195 


2.90 3.58 


09 


1355 


24.9 772 



But when we compare the average carbon dioxide production and 
oxygen consumption per kgm. per minute of the men with that 
of the women, there is a striking agreement in the results. Thus, 
3.04 cc. of carbon dioxide per kgm. of body weight were pro- 
duced per minute by the men against 2.90 cc. for the women, and 
3.65 cc, of oxygen were consumed by the men as compared with 
3.58 cc. of oxygen by the women. In comparing the calculated 
heat production per twenty-four hours, the greater total amount 
for the men is obviously due to the larger body weight; but on the 
basis of per kgm. of body weight, 25.5 calories were produced 
with the men as compared with 24.9 calories with the women; 
per square meter of body surface the values are 832 calories for 
the men and 772 calories for the women. 

'"Benedict and Smithi thie Journal, xx, p. 243, 1915. 



F. G. Benedict and L. E. Emmes 257 

A comparison of this nature is, however, distinctly erroneous, 
and while one might assume that with eighty-nine men and 
sixty-eight women the average figure might hold, nevertheless 
a careful analysis of the situation will show that this method of 
comparison is not wholly justifiable. Indeed, its use has been 
carefully avoided and rightly so by all previous writers. On the 
other hand, believing that only individuals of like weight and 
height can properly be compared, it is possible with such a large 
accumulation of material as is presented in the two tables pre- 
viously published to compare the metabohsm of women with that 
of men of approximately the same weight and height. Such a 
comparison is made in Table II. Furthermore, although dis- 
inclined to believe in the value of a comparison on the basis of 
heat production per kgm. and per square meter of body surface', 
we have included these comparisons in the table, since they have 
been used extensively by other writers. 

The attempt in this comparison has been to secure a fair aver- 
age picture of the metabolism of men and w^omen, and to select 
the data for the various groups of men and women compared 
so as to keep the difference in height within 1 or 2 cm. and the 
difference in weight within 1 or 2 kgm. Unfortunately, in all 
instances this method of comparison cannot be adhered to, and 
certain deviations are absolutely necessary. While for the 
most part it has been possible to employ three or more men and 
women in each group, with certain groups, particularly Groups 
V, VI, VII, X, and XI, it has been impracticable to use more 
than two individuals. This is admittedly a failing in the method 
of comparison. 

In both Groups I and II all the women had a lower metabolism 
per kgm. of body weight and per square meter of body surface 
than any of the men with whom they are compared. In these two 
groups, therefore, the evidence is strongly in favor of the general 
supposition that the metabolism of women is less, not only 
absolutely, but per kgm. and per square meter, than that of men 
of similar weight and height. In the subsequent groups the 
differences are not so noticeable until we reach Group V with 
unfortunately only one woman for comparison. In Group VII 
but one man is compared, but his metabolism is distinctly 
higher than that found with normal men of approximately his 



258 Metabolism of Normal Men and Women 



TABLE II. 



Comparison of the heat production of normal men and women of like Uodij 

weight and height. 



QROUP AN'D 8UBJKCT 



NX7DE 
WEIGHT 



kgm. 



GrOI'P I 

Men 

F.P.H 65.1 

K. H. T 64.7 

D. M 64.0 

Al. J. ^ 63 .7 

R.G 62.7 

H. C. B 62.0 

Women 

MissC. Z 67.2 

Miss S 65.6 

Miss A. K 63.2 

Group II 

Men 

jM. J . S 63 .7 

S. A. K 60.8 

IT. B. H 60.5 

W. F. B 60.1 

T. H. Y 59.2 

Women 

MissC. H 63.4 

Miss \. \ 62.9 

MissC^ 61.9 

Groip III 

Men 

H. H. A 62.3 

S. A. R 60.8 

W. B. L 59.3 

Women 

Dr. A. B 60.3 

Miss B. W 59.4 

Mrs. E. B 58.0 

Miss M. M 57.9 

Group IV 

Men 

A. L 60.6 

H. B. R 60.5 

J. B. T 60.1 

W. F. B 60.1 



HEIGHT I 



cm. 



heat pkoduction per 24 hkh. 
(computed) 



Total Per kjin. ' ^^"^ «*}""^ 
* meter 



cal. 



cal. 



:al. 



173 


1.543 


23.7 


775 


170 


1499 


23.2 


757 


171 


1651 


25.8 


83S 


170 


1647 


25.9 


840 


173 


1590 


25.4 


820 


173 


1653 


26.7 


85() 


170 


1521 


22.7 


747 


171 


1426 


21.8 


713 


171 


1402 


22.2 


717 


170 


1499 


23.2 


757 


170 


1647 


25.9 1 


840 


165 


1460 


24.0 


768 


168 


1487 


24.6 


783 


168 


1632 


27.2 


863 


169 


1605 


27.2 


861 


166 


1413 


22.3 


722 


168 


1330 


21.2 


682 


168 


1427 


23.1 , 


739 


164 


1487 


23.9 


770 


165 


1460 


24.0 


768 


164 


1451 


24.5 


776 


163 


1486 ' 


24.6 1 


786 


162 


1546 


26.0 ' 


827 


163 


1415 


24.4 ' 


769 


164 


1475 


25.4 


801 


171 


1576 


26.0 


829 


168 


1487 


24.6 


783 


171 


1748 


29.1 


925 


168 


1632 


27.2 


863 



F. G. Benedict and L. E. Emmes 



259 



TABLE II~Coniinued. 



OXOUP AND SUBJECT 



NX7DX 
WKXQRT 



kgm. 



BSIGBT 



» HEAT PBODUCTXON FXB 24 BBS. 
(COMFDTXD) X 



cm. 



Total 



cal. 



Per kcm. 



eal. 



I 



Per square 
meter 

eal. 



(Jroup IV — Continued 












Men 












T TT V 


59.2 


169 


1605 


27.2 


861 


F T W 


57.8 
57.2 


169 
167 


1472 
1616 


25.5 
28.3 


800 




883 


^^^^1 A ^L^^^ M ^K^L AaAA^A^K ^ A A A a A ^ 


57.1 
57.0 


171 
169 


1539 
1531 


27.0 
26.9 


844 




841 


Women 




Miss L. U 


59.3 
58.6 
58.1 


L69 
167 
168 


1448 
1429 
1518 


24.4 
24.4 
26.2 


774 


MissM. W 


768 


MissM. P 


823 


C.ROUP V 




Men 












LEE 


59.8 


175 


1707 


28.5 


908 


D. J. M 


58.0 
56.1 


175 
173 


1615 
1522 


27.8 
27.1 


878 


V • ^^e V^« ••••••••■■•■•• 


846 


V/ovMn 












MissE. P 


57.7 


175 


1430 


24.9 


780 


Group VI 












Men 












p -p 1 


57.2 


167 


1616 


28.3 


883 


X^l • X^ • aV« •••••••••••»• 


55.2 


164 


1341 


24.3 


753 


WwMn 












MissL. K 


56.8 


166 


1365 


24.1 


750 


MissG. L 


55.0 


166 


1480 


27.0 


832 


Group VII 












Men 












T A -p 


54.9 


156 


1612 


29.4 


906 


FKomen 












ivirs. 1./. \j 


54.9 


153 


1276 


23.3 


719 


MissL. T 


53.6 
52.2 


155 
158 


1247 
1415 


23.3 
27.2 


713 


Miss B 


823 


Group VIII 




Men 












Dr. P. R 


55.2 


164 


1341 


24.3 


753 


M. B 


53.6 


160 


145.5 


27.1 


831 


Wovfien 












Miss J. C 


55.1 • 


162 


1363 


24.8 


764 


MissM. T 


54.5 


164 


1359 


25.0 


770 


MissF. K 


54.1 


164 


1262 


. 23,4 


716 


MissF. E 


53.1 


162 


1391 


26.2 


799 



26o Metabolism of Normal Men and Women 



TABI^ n— Concluded. 



(;KOnP AND SUBJRCT 



XUDE 
WEIGHT 



kgm. 

Group IX 

Afen 

J. J. G 50.2 

V.E.H '■ 49.3 

T. M. C 48.5 

Women 

Miss J. B 51.1 

MissE. C 50.5 

Miss I. B 50.1 

MissC. B 49.8 

Group X 

Men 
T. M. C 48.5 

Women 

MissL. B 47.13 

MissE. T 46.7 

Group XI 

Men 
J. H 46.3 

Women 

MissR. W 45.0 

MissM. J 44.8 



HEIGHT 



cm. 



hkat production feb 24 hr8. 
(computed) 

Total • Per kgm. Po^X?" 



cat. 



cat. 



cal. 



164 


1425 


28.4 ■ 


848 


163 


1365 


27.7 


822 


165 


1292 


26.6 


788 


163 


1265 


24.8 


746 


164 


1327 


26.3 


788 


166 


1233 


24.7 


737 


162 


1419 


28.6 


850 



165 

167 
164 



154 

153 
157 



1292 

1168 
1336 



1223 

1273 
1189 



26.6 

24.9 , 
28.6 



26.4 

28.3 
26.5 



788 

730 

838 



769 

816 
767 



height and weight. In Group IX the general picture would imply 
a metabolism somewhat higher with the men than with the women. 
It is obvious that the best comparison for all these groups can 
only be made with the average figures; we have therefore sum- 
marized the comparisons for the several groups in Table III. 
From the conditions of the grouping it is obviously unnecessary 
to consider the total metabolism, and our comparisons may simply 
be confined to the heat production per kgm. and per square 
meter of body surface. In these eleven groups the men show a 
greater metabolism per kgm. of body weight than that of the 
women in eight cases; the average of the data shows that the men 
had a heat production of 26.5 calories per kgm. per twenty-four 
hours as compared wnth a heat production of 25.0 calories per kgm. 
for the women, or approximately a 5 per cent increase. Further- 



F. G. Benedict and L. E. Emmes 



261 



more, on the basis of per square meter of body surface, we find a 
greater heat production with men in nine cases, the average be- 
ing 819 calorics per square met<>r for the men as compared with 
770 calories for the women, an increment of approximately 
G per cent. 

Some of the defects of this method of comparison are obvious 
when we consider the special groups. Thus, in Group I the men 
arc* of approximately normal body weight while the women with 

TABLE III. 

Summary of comparison of normal men and women of like body weight and 

height. 





1 




hkat pboduction pkr 24 ubs. 
(computed) 






GROUP NO. 

1 


Per kgiu. of body woisht 


Per square nietei 
surface 

Men 

1 


■ of body 




Men 


1 

Women 


Women 






cal. 


eal. 

1 


eal. 




cal. 


I.. 




25.1 


i>2.2 


814 


1 


726 


II.. 




25.4 
24.1 
26.9 
27.8 


22.2 
25.1 
25.0 
24.9 


812 

771 

o'xo 

877 


i 

1 

1 


714 


Ill 


796 


IV 


788 


v.. 




780 


VI. 




26.3 


25.6 


818 




791 


VII. 




29.4 
25.7 


24.6 
24.9 


906 
792 




752 


VIII. 




762 


IX. 




27.6 


26.1 


819 




780 


X. 




26.6 


26.8 


788 




784 


XI. 




26.4 


27.4 


769 




792 


Average 


26.5 


1 25.0 


819 


._. 1 ._ 


770 



whom they are compared are larger than nonnal, and in this group 
wo find a greater disproportion between men and women than in 
any of the others. On the other hand, in Groups X and XI, in 
which small individuals arc compared, the metabplism is slightly 
higher with women than with men; if we are to explain this on 
the basis of an actual difference in the active mass of protoplasmic 
tissue, we must infer that with small, thin women the proportion 
of subcutaneous fat is no greater than that with small, thin men. 
It would appear from these comparisons, therefore, that the 



262 Metabolism of Normal Men and Women 

metabolism of men is about 5 or 6 per cent greater than that of 
women with like height and body weight. Since athletes have 
been carefully excluded m this selection of material, we deal here 
only with approximately normal individuals, and in any event 
individuals of similar size and form appear in the comparative 
groups. The slight increment in the metabolism of men over 
women may be explained on several grounds. First, there may 
be a disproportion between body weight and body surface; in 
other words, the Meeh formula may not hold with women as it 
does with men. Doubtless in individual cases this may play a 
part; but with as large a number of individuals as are here studied, 
we think it fair to assume that this disproportion is more theoreti- 
cal than actual. Secondly, there is in all probability with women, 
particularly in the groups with the greater body weight, a larger 
proportion of subcutaneous fat than with men, and conversely 
with men a larger proportion of active protoplasmic tissue. 
Finally, it should be stated that while the age of certain individ- 
uals in the comparison should properly be considered, the greater 
number of our experiments were made with individuals between 
twenty and thirty years of age, aud hence no individual case can 
in any way aflfect the general deduction; i.e., that men have a 
slightly greater metabolism per kgm. of body weight and per 
square meter of body surface than have women of like weight and 
height. 



warn 



FACTORS AFFECTING BASAL METABOLISM. 

Bt FRANCIS G. BENEDICT. 

{From the Nutrition Laboratory of the Carnegie Institution of Washington, 

Boston.) 

(Received for publication, January 15, 1915.) 

Superficial observations have led us to the general belief that 
the vital processes are much the same with average individuals. 
It is true that we are inclined to classify people as high-strung 
and nervous on the one hand and phlegmatic on the other; yet, 
on the average, the normal individual apparently has a meta- 
bolic plane comparable to his neighbor's, for the general uni- 
formity in the energy requirement, as indicated by dietary 
studies of people of various classes performing like amounts of 
muscular work, suggests that there is likewise a uniformity in 
the basal metabolism of these individuals, t.e,, the metabolism 
during complete muscular rest and in the post-absorptive con- 
dition. Undoubtedly a group of miners will have about the 
same average metabolism as another group of miners, or a group 
of stenographers in one city will have much the same energy 
requirement as a group of stenographers in another city. But 
while the uniformity in energy requirement is of great signifi- 
cance in dietary studies and in the arrangement of dietaries, 
from the standpoint of pure physiology and for the application 
of physiological values to a study of pathological cases, it is of 
much more importance for us to know how constant is the basal 
metabolism with different individuals. For instance, while 
two groups of stenographers would have on the average the 
same energy requirement for the like amount of work, would 
the individual members of the group have the same basal 
metabolism? 

The general impression that individuals of like occupation 
have similar dietary needs applies only to persons of approxi- 
mately similar age and weight; for it is a matter of household 
observation that the growing, active child apparently requires 

263 



264 Factors Affecting Basal Metabolism 

a much larger proportion of food than does the more sedate 
adult. Furthermore, it is conunonly supposed that a large 
man would have a much greater basal metabolism than would 
a small man. Hence the factor of size should be considered, 
and it is clear that only individuals of like size should be com- 
pared. Heretofore the chief factor in determining the size of 
individuals has been the weight, and many physiologists, as a 
closer approximation to a basis for scientific comparison, have 
been wont to express values on the basis of the metabolism per 
kilogi'am of body weight. 

Other physiologists use another index: that of the metab- 
olism per square meter of body surface. They base their 
hypothesis upon one of the two current explanations of heat 
production in the body: first, that heat is incidental to the 
vital processes, all the energy used by the body being finally 
degraded to heat which, as the end-product, is lost from the 
body; and, second, that heat is produced primarily to keep the 
body warm. 

Of these two views, the basal idea in the second theory has 
obtained for many years. It is a well known fact that in health 
the normal body temperature of approximately 37°C. remains 
relatively constant throughout life. The body is continually 
losing heat; since the body must be kept at a normal tempera- 
ture, heat must be produced to supply the heat lost. The 
production of this heat may take place in two ways: first, by 
active external muscular work, such as would be exemplified 
by teamsters swinging their arms in cold weather to keep warm; 
second, by internal muscular activity, such as increased blood 
flow, muscular tonus, etc. 

The development of the belief that the heat is produced pri- 
marily to keep the body warm may easily be traced from the 
old conception of the animal organism as a cooling body to which 
Newton^s law of cooling applies. As early as 1843, Bergmann^ 
promulgated this idea, although he gave no experimental evi- 
dence. Forty years later Rubner,^ in his classical research on 

' C\ Bcrgmann and R. Leuckart: Anatomisch-phyaiologiches Vbersicht 
des Thierrrichs, Stuttgart, 1852, p. 272. See also Berginann: W&rmeokonO' 
mic der Thiere, (iottingon, 1848, p. 9. 

^M. Rubner: Ztachr. f. BioL, xix, p. 545, 1883. 



F. G. Benedict 265 

animals, and simultaneously Richet,^ brought forth evidence 
to signify that the heat production of living animals was directly 
proportional to the surface of the body, thus apparently sub- 
stantiating in every way the belief that the animal organism 
acted as a cooling body, and heat was produced to maintain 
the body temperature. Indeed, Rubner, citing his measure- 
ments of the heat production of several groups of animals, 
maintained that approximately 1000 calories of heat per square 
meter of body surface per twenty-four hours were given off by 
the living animal. Later Voit* concluded that the law of the 
constancy of heat production per square meter extended over a 
much wider range in the animal kingdom than had at first 
been believed. 

This view of the intimate relationship between the surface 
area of the body and its heat requirement has had a strong hold 
upon European and American scientists for many years. This 
fact is emphasized by the regularity with which scientific re- 
search is presented in which the heat production is considered 
on the basis of per square meter of body surface. While the 
experimental evidence upon which this law was based was ob- 
tained exclusively in observations with animals, sufficient experi- 
mental evidence has been accumulated with human beings of 
varying ages, sex, and activities, to permit a critical discussion 
of the metabolism and the various factors affecting it to find 
if the laws of heat production per unit of body weight and of 
body surface are scientifically sound and if they obtain with the 
human organism. 

^ While the factor of body size is immediately recognized by 
all physiologists as a variant in studying basal metabolism, 
other possible factors arc sex, age, muscular development, and 
the character of the preceding diet. In the seven years since 
the establishment of the Nutrition Laboratory, a definite in- 
vestigation has been in progress to secure values for the basal 
metabolism of a large number of individuals of all ages for an 
ultimate analysis of causes and laws governing the heat lost from 
the body. Therefore in this discussion all other researches are 
disregarded simply because they were not made primarily with 

' C\ Richct: Arch, dc physiol.y xvii, p. 284, 1885. 
^E. Volt: Ztschr. /. BioL, xli, p. 120. 1901. 



266 Factors Affecting Basal Metabolism 

a view to securing data of this kind, and it thus seems wisest 
and most logical to consider only those values which were se- 
cured primarily for this purpose with as nearly as possible like 
conditions, with similar technique, and with due regard to all 
the extraneous external factors known at the time to affect 
metabolism. 

The data at present consist of observations upon eighty-nine 
men and sixty-eight women,* all of whom may be grouped as 
normal individuals, meaning thereby that they are people in 
apparently good health. In addition we have data for a large 
number of new-born infants* and a few infants under one year 
of age who would be distinctly classed as normal. A consider- 
able number of observations are also available which were made 
upon infants who, though atrophic, were not otherwise abnormal." 

In the studies with adults, the two important factors affecting 
metabolism, namely, the influence of the ingestion of food and 
the influence of external muscular activity, were completely elimi- 
nated, the first by studying the subject only in the post-absorp- 
tive condition — that is, at least twelve hours after the last food — 
and the second by a graphic demonstration of complete muscular 
ix^pose. This graphic record was obtained by one of the various 
forms of technique developed in this lal)oratory, in which the 
subject is either provided with one or two pneumographs about 
the chest and thighs, or lies upon a bed which records the slightest 
alterations in the center of gravity of the body. Our main 
problem, then, is to find what factors affect the quiet, resting 
metabolism, when the subject is in the post-absorptive condition. 

Total metabolism compared with total body weight. 

We may examine, first, the relationship between total metab- 
olism and the body weight. Since we should expect that a 
large individual would produce more heat than a small individual, 

* F. G. Benedict, L. E. Enimes, P. Roth, and H. M. Smith: this Journal, 
xviii, p. 139, 1914. The recently appearing results of W. W. Palmer, J. H. 
Means, and J. L. Gamble {ihid.^ xix, p. 239, 1914) are accorded special 
notice. 

• F. G. Benedict and V. B. Talbot: Am. Jour. Dis. Child., viii, p. 1, 1914. 
^Benedict and Talbot: Carnegie Institution of WashingUm Publica- 
tions, No. 201, 1914. 



F. G. Benedict 267 

it is of interest to see if a definite relationship exists between 
the body size and the total metabolism. Consequently, with 
the conmionly used index of body weight as an index of size, 
the heat production of our subjects, both male and female, may 
be studied on the basis of the total heat production compared to 
the total weight. For this comparison it has been deemed most 
advantageous to plot the values for the eighty-nine men and 
sixty-eight women in the form of charts with the ordinates as 
total calories and the abscissae as body weight. These are given 
for the men in Chart I and for the women in Chart II. By 
thus separating the men from the women we take immediate 
cognizance of the fact that the two organisms may not advan- 
tageously be compared, as shown in the foregoing paper by 
Henedict and Emnies.^ Likewise, in order to emphasize tho 
fact that athletes* as a class may not Ix? properly compared with 
non-athletes, the athletes on Chart I are marked by small crosses 
instead of dots. Finally, for purposes of comparison, wo have 
added to these charts the values recently published by Palmer, 
Means, and Gamble^^ for a group of normal young men and 
normal young women, designating them by dots enclosed in 
circles. Special discussion of the athletes and the normal cases 
of Palmer, Means, and Gamble will enter into the consideration 
of all these data. 

An examination of C'hart I shows that, as would be expected, 
the men of small body weight have generally a much less total 
heat production than those of large body weight, and yet the 
relationship is by no means a clear one. Thus, of the numerous 
subjects with a total energy output of approximately 1600 calories 
[)er day, one at least has a body weight of less than 50 kgm., 
and another has a body weight close to 83 kgm., showing two 
organisms of widely varying weight producing essentially the 
same amount of heat per twenty-four hours. 

It is of interest also to note the possible variations i^i the heat 
output of various individuals of the same body weight. Thus, 
with those weighing about 58 to 60 kgm. we have a variation 
from 1331 calories to 1748 calories, and with those weighing 

• F. G. Benedict and L. E. Emmes: this Journal, xx, p. 253, 191.5. 
» F. G. Benedict and H. M. Smith: ibid., xx. p. 243, 1915. 
^'Palmer, Means, and Gamble: loc. rit. 



268 Factors Affecting Basal Metabolism 

about 80 kgin.. the variation is from 1615 calories to 2126 calories, 
these representing the extremes. It is clear from Chart I that 
the athletes as a class have a very much higher metabolism than 
havp the normal individuals of the same weight. 



C»LS 




p 


[ 




r 








las 


3fl- 




































































' 


- 


- 




































■ 
















.■ -1 




,™ 










■ ■/ ! 


































• 1 


-• 




, 








im 










. 












' 






7^ 






















J 


■ 








































uno 




















































































































IISO 

























(.'hart I. t'ompariaon of body weight and total twenty-tour hour 

heat jirodiiption of iiincty-aeven normal men. 



In the observations of Palmer, Means, and Gamble, presum- 
ably great care was exercised to secure the basal metabolism of 
the individuals studied, and there is eveiy indication that most 
careful attention was given to experimental technique. Their 



F. G. Benedict 



269 



values may therefore lie taken as those which would be reason- 
ably expected from a group of normal individuals. It will be 
seen that even in this small, selected group there was a consider- 
able divergence in the heat production, and that it does not 
appear to be a function of the body weight. 



CALS 
































































































" 






























































































., 


A 


*• 




















































I 






















■. . 


















































.- 




■ 




































































































• 





















60 "iS 70 



75 80 65 90 95 

iml total twenty-four hour 



Chakt II. (.'ompariHon of bo<ly w 
heat prtxluirtion of Hcvonlj-fcven normiil woiiii'n. 

If we consider the normal values for women wliich arc jjlolted 
in Chsu^ II, we find here also a general tendency for the heavier 
individuals to have the larger nietabolism, although there is by 
no means a regularity in the relationship between l)ody weight 
and the total metabolism. Tlius, two women, having a total 



270 Factors Affecting Basal Metabolism 

heat output of 1475 calories, have as a matter of fact a body 
weight ranging from 48 to 80 kgm. Perhaps the most striking 
difference is that between two individuals having a body weight 
of approximately 60 kgm., the total heat production being re- 
spectively 1187 and 1666 calories. When we consider by them- 
selves the values obtained by Palmer, Means, and Gamble, 
designated here also by dots enclosed in circles, we note like- 
wise that individuals having approximately the same heat out- 
put, namely, 1450 calories, may have a difference of 27.5 kgm. in 
weight, with no approximation to a constant relationship between 
the total metabolism and the total body weight. 

It is clear from an examination of Charts I and II that the 
relationship with both men and women between total body 
weight and total heat production is therefore only a most general 
one and the diversities in the values found are so great as to 
make it impossible to establish anything approximating a uni- 
formity; thus we find here not the slightest evidence of a law 
governing the relationship between the total body weight and 
the total heat production. 

Heat production per kilogram of body loeight. 

In the foregoing discussion the common index to size, namely, 
weight, has been utilized, the assumption being made that in this 
form of comparison there is tacit assent to the general conception 
that mere weight is the determining factor. It was early recog- 
nized, however, that it is somewhat illogical to compare tlu» 
total heat output of two individuals varying widely in size, as, 
for instance, a small infant with that of an adult, and one of 
the earliest attempts to secure uniformity aijd a rational basis 
of comparison was the expression of the total metabolism in 
terms of heat per kgm. per twenty-four hours.^^ The extensive 
use of the heat production per kgm. of body weight by Conti- 
nental and American writers makes it necessary for us to examine 
very closely the propriety of this procedure. The values for the 
metabolism measurements obtained- with the eighty-nine men 

** J. Forster: Ayntl. Ber. d. 60 Versammlung deutsch, Naturforscher u. 
Aertze in MuncheUj Munich, 1877, p. 355. Forster used 10 kgm. of body 
substance as Iuh unit. 



F. G. Benedict 



271 



and abcty-cight women previously refcripiil to have therefore been 
[)lotted on this basis in Charts III and IV: the heat production 
per kgm. of body weight is given in the ordinates, and the body 
weight in l^m. in the abscissae. 

The entire absence of uniformity with the men subjects is 
even more strikingly shown in Chart III than in Chart I. While 
there may lie a slight tendency for the individual of the smallest 





















































































































































'■■ 


























































■J 




. ' 


























\ 


■ 
































































































































„ 





























6S 70 

Cbart III. Comparison of body weight and heat production tier 
ktcm. of body weight per twenty-four hours of ninety-seven normal men. 



Ijody weight to have the highest licat production per kgm. of 
lx)dy weight, on the other hand we have at least two individuals 
(vith a heat production of about 23.5 calories per kgra. of body 
weight, one of whom weighs 108 kgm. and the other 50 kgm., or 
less than one-half the weight of the first subject. Furthermore, 
of two individuals having very nearly the same body weight (50 
kgm.), one shows a heat production of 23.2 calorics per kgm. 
of body weight and the other 32.3 calories per kgm. of body weight. 



272 Factors Affecting Basal Metabolism 

In other words, there is no possibility of finding a gcm-ial rplii- 
tionship between body weight and the heat production pci- kgni. 
of body weight. 

The values for the athletes show a tendency to group towards 
the upper side of the ehart. Of special interest arc the added 
yahies for the subjects of Palmer, Means, and Gamble, foi- even 

















































































































































• 






















: 


-. 
















































































,- 






















" 


f 








, 






















* 




























• 


























































* 






























































KC 


IS. 41 


) 4, 


5 51 


5 


5 6 


fl 


S 7 







5 e 


G 


S 'i 





, <'haht IV. Compuriaon of body weight anil heat production per liem, 
of body woiKlit per twenty-four hours of aoventy-Beven normal women. 

these selected normal individuals of 20 to 30 years of age have 
a heat per kgm. of body weight ranging from 21.0 to 25.4 calorics, 
with no tendency towards a constant relationship. 

With the women subjects shown in Chart IV, the grouping 
is much the sjunc as for the men subjects in Chart III. Although 
the values for the heat production per kgm. of Ijody weight range 
from IS. I calorii'M to 32.1 calories, and, in general, the women 



F. G. Benedict 273 

with the smallest body weight show a tendency to have the 
highest heat production per kgin. of body weight, there is neverthe- 
less an absence of uniformity, particularly with the women whose 
body weights range from 50 to 60 kgm. The tendency towards 
a constant relationship is, however, much more evident than 
with the male subjects.. 

Of the comparisons thus far noted, it is probable that the gi'oup 
of women subjects studied by Palmer, Means, and Gamble come 
the nearest to approximating regularity; for while we find actual 
variations on the basis of per kgm. of body weight of 18.6 calories 
to 27 calories, nevertheless a probable line drawn through these 
points would imply that if the heat production per kgm. of body 
weight as well as the actual body weight were taken into con- 
sideration, there would be a semblance to regularity. Project- 
ing the most probable straight line through these points, we find 
variations of plus or minus 8 to 10 per cent, variations still too 
great to permit this relationship to be classed as a physiological 
law. 

With men we have found that the heat production per kgm. 
of body weight ranged from 19.7 calories to 32.3 calories, 2.6., 
a range of 60 per cent or more; with women from 18.1 Calories to 
32.1 calories, or a range of 80 per cent. With infants Benedict 
and Talbot found that the values varied from 42 to 88 calories, 
including both normal and atrophic infants; with the normal 
infants alone, they found a range in the heat production from 
42 to 62 calories, or a difference of over 40 per cent. It is obvious, 
therefore, that any system of comparison which results in an 
(»rror of 60 per cent with normal men, 80 per cent with normal 
women, and 40 per cent with normal infants cannot be con- 
sidered as having the fundamental characteristics of a law, and 
that accordingly this system of notation should be discarded. 
If we apply this unit of measurement to the specially selected 
normal individuals of Palmer, Means, and Gamble, we find a 
variation of 21 per cent with the men and 45 per cent with the 
women. It is thus clear that the metaboUsm per kgm. of body 
weight has no claim to be considered as a physiological func- 
tion, and its usage is accordingly no longer justified. 



2 74 Factors Affecting Basal Metabolism 

Heat per unit of body surface. 

We have now to consider the second common unit for measure- 
ment of the heat production; namely, the heat production per 
unit of body surface. This, while fundamentally based upon 
Newton's law of cooling, and of the similarity in body surface of 
similar geometrical solids, actually calls for an elaborate study 
of the measurement of the body surface of man to establish a con- 
stant to be used in computation. It is known that the surface of 
irregular but similarly shaped bodies varies as the cube root of the 
square of the volume. By measuring a number of individuals. 
Meeh found that with adult man a constant of 12.312 should be em- 
ployed in using this formula. Subsequent to the appearance of 
Meeh's paper," his constant has been by common consent employed 
by numerous physiologists for computing the body surface of adults. 
The constant for infants was found by Meeh to be 11.9, although 
subsequent measurements by Lissauer^' have shown that the 
value 10.3 is more nearly accurate. For the measurement of th(* 
metabolism of adults, the Meeh formula has been commonly 
used. Doubt has been thrown, however, upon the accuracy 
of this formula, and in a number of instances when a disagree- 
ment between Rubner's law of the constancy of the heat produc- 
tion per square meter of body surface has been observed, the 
popular method of explaining such irregularities has been to 
assume a disproportion between the body surface and the weight 
and an erroneous application of the Meeh formula. Rubner'^ 
himself has considered this with two boys gf widely var^dng body 
composition, and showed a possible variation in the formula 
of about 7 per cent. 

Employing the Meeh formula we have computed the body 
surface of all of our men and women subjects and plotted the 
values found for the heat production per square meter of body 
surface in Charts V and VI. With our men (Chart V) the values 
ranged from 693 calories per square meter of body surface p(»r 
twenty-four hours to as high as 958 calories per square meter of 
body surface. In examining this chart it should be taken into 
consideration that according to the law which has been used by 

^2K. Meeh: Ztschr. f. Biol., xv, p. 425, 1879. 
" W. Lissauer: Jahrb. /. Kinderheilk.j Iviii, p. 392, 1903. 
^*Rubner: Beitrdgc zur Ernahrung im Knabenalter mit besondtrer 
Beriicksichtigung der Feitsucht^ Berlin, 1902, p. 40. 



F. G. Benedict 



275 



physiologists for many years as a basis for comparison, all the 
points would be expected to fall in a horizontal line. Even 
with the selected subjects of Palmer, Means, and Gamble, the 
values range from 726 calories to 818 calories, a variation — ap- 



940 


• 








• 












920 




• 
• 


• • 


X 

» 














900 




4 


• 4 












911 C 

loae 


KOsi'* 


880 






• 


• 




X 










860 






4 
»x . 


• 




• 




X 






840 




» 


• 


• X 

• • 

> 


X 


K 

X 


X 

• 


X 






820 


• 


• 
• 


X 


• 

X 


• 


• 






X 




800 






( 

• 




• 


• 
• 


• 








780 


• 


i 


• 


i 

• • 


• 


• 








• 


760 


• 




• 


1 

• • 

• 


► 


• 


► 


• 


• 




740 






» • 


• 


• 












720 






• 






• 


• 








700 






• 






• 




• 






680 


1 


» 












• 







KGS. 50 55 60 65 70 75 80 85 90 95 

Chart V. Comparison of body weight and heat production per square 
meter of body surface per twenty-four hours of ninety-seven normal men. 

proximately 12 per cent — which would hardly designate this 
supposed relationship as a physiological constant. 

In a recent paper Du Bois*^ has maintained that an average 

" E. F. Du Bois: Jour. Am, Med. Assn., Ixiii, p. 827, 1914. In this 
connection attention should be called to the attempts of D. and E. F. 
Du Bois to improve upon the method of computing the body surface by 
a series of carefully selected moasuremcnts (D. and E. F. Du Bois: Arch. 
Int. Med.y xv, (in press) 1915). 



3^6 Factors Affecting Basal Metabolism 

of a selected nuinlwr of normalg, fixnn which he has excluded 
athletes (a debatable procedure) and all individuals other than 
those of perfectly normal proportions, has given an average 
value of 34.2 calories as the heat production per square meter 
per hour, or 821 calories per square meter per twenty-four hours. 
A large number of our so called "normal" subjects would l>e 



tMS. 








. 
















































' 










































































































-_ 


., 




























^- 




























































. 














," 


-■ 






•• 
























































* 





























































KG3. *0 t5 50 55 60 65 

Chart VI. Cumparison of body weight und hcnt production |ior square 
meter itf body tmrfnee per twonty-tour hours of seventy-seven normal 



excluded by Du Hois from his classiheatton on account of their 
being athletes or people of unusual shape, but the values found 
by Palmer, Means, and Oaiiible were for the most part secured 
with individuals of such size and weight as would fully conform 
to Du Hois's specifications. Yet wc find that all the values given 
by Palmer, Means, and (Iambic fall Iwlow the Du Boia figure; 



F. G. Benedict 277 

indeed, one value falls 10 per cent below. We also find that 
all the athletes among our subjects are without exception above 
the value of 821 calories, although but two are more than 10 
per cent above this value. 

Examining the plotted values for women (Chart VI) we find 
the range to be from 633 calorics per square meter to as high as 
9G5 calories per square meter, with no tendency for the values 
to fall in a horizontal direction or to average at any given value. 
With the normal women of Palmer, Means, and Gamble, we 
likewise find an extensive range, namely, from 650 calories to 
843 calories per square meter of body surface, with no approach 
to constancy in the arrangement of values. 

The gross irregularities in these plots agree wholly with the 
variations found for infants by Benedict and Talbot. Includ- 
ing all the infants studied, the values for the heat production 
per square meter of body surface ranged from 554 calories to 1334 
calories; the strictly normal infants varied from 554 calories to 
998 calories. 

It is obvious that any basis of comparison which involves pos- 
sible variations of 40 per cent with men, of 43 per cent with 
women, and 80 per cent with normal infants, cannot l>e con- 
sidered as a physiological law. If a special selection is made, 
as was done by Du Bois, it may be possible so to select a certain 
group as to approximate constancy. It seems particularly 
unfortunate, however, that the normal individuals who are of 
the greatest value in compai'isons with pathological cases are 
not those of normal height and weight. The majority of path- 
ological individuals are either above or below weight, and hence 
it is necessary to use for comparison only normal individuals 
who lie outside of the supposed regular or normal shape of indi- 
viduals^ For instance, it is obviously unfair to compare an 
emaciated diabetic^^ with a normal, well nourished man. A man 
who was thin and of the same height and weight as the ema- 
ciated diabetic would, by the process of selection above referred 
to, be excluded from a normal grouping as being impossible to 
use. Even the selected groups studied by Palmer, Means, and 
Gamble show difTerences between the lowest and highest values 
of about 12 per cent for the men and 30 per cent for the women. 

" F. G. Benedict and K. P. Joslin: Carnegie Institution of Washington 
Fublicaiions, No. 170, p. 120, Tables 131, 1912. 



278 Factors Affecting Basal Metabolism 

If, however, we are to consider that surface area of the body 
determines the resting metabohsm, we have no right to apply 
this law solely to a selected group of normal individuals. It 
must, if it is a physiological law, apply as well to a new-born 
infant, to an atrophic infant, or to an athlete. Large varia- 
tions in the heat production per square meter of body surface 
are noted with these groups, variations which can in no wise 
be explained, by the widest assumption of a discrepancy in the 
relationship between the body weight and the body surface or of 
an error in the conmionly used formulae for this computation. 
It must furthermore be recognized that while, with select<^d 
normal individuals, a constancy approximating plus or minus 
10 per cent may be said to obtain, there is grave danger in laying 
great stress upon this apparent agreement; for it may be con- 
sidered as a fundamental tenet of physiological experimentation 
that, even among normal individuals. Subject A may not be 
compared to Subject B even if the two are of the same height 
and weight. On the other hand, in studying disturbances of 
metabolism as a result of disease or of a superimposed factor, 
it is perfectly permissible to compare a group of fifteen or more 
persons with an equal number of normal individuals of the same 
height and weight; for, in such a group comparison, the same 
errors which may enter into the application of a formula for com- 
puting body surface from body weight would obtain with the 
two sets of individuals compared. An excellent illustration 
of this may be found in the observations in this laboratory on 
diabetics in which, by applying the group system, an increment 
in the basal metabolism amounting to approximately 20 per 
cent was noted in severe diabetes. 

That there is with normal individuals a rough approximation 
between basal metabohsm and body surface is not surprising 
in view of the recent investigations on growth.^^ If the blood 
volume, area of the trachea, and area of the aorta have been found 
to bear a simple mathematical relationship to the total body 
weight in normal individuals — a relationship expressed by the 
cube root of the square of the body weight — it is not unlikely 

1' G. Dreyer and W. Ray: Phil, Trans., Series B, cci, p. 133, 1909-10. 
G. Dreyer, W. Raj', and E. W. A. Walker: Proc. Roy. Soc, Series B, 
Ixxxvi, pp. 39 and 56, 1912 and 1913. 



F. G. Benedict 279 

that the active protoplasmic tissue of the body, z.e., the mass 
of heat-producing matter, as well as the surface will have 
approximately the same mathematical relationship. Further- 
more, if the heat production is proportional to the cube root of 
the square of the weight, it is due not to the fact that surface 
area determines thermolysis, but that the mass of active proto- 
plasmic tissue, with probably the same mathematical relation- 
ship to the body weight, determines thermogenesis. 

It is clear, therefore, that even with normal individuals a re- 
lationship between body surface and heat production which may 
be expressed with any approximation to mathematical accuracy 
does not exist. Hence, in considering the metabolism of nonnal 
men and women, we are compelled to maintain that the so called 
*'law*' of heat production per square meter of body surface does 
not obtain. 

Relationship between heat prodxiction and body composition. 

The two laws supposed to determine the metabolism in the 
human individual, i.e., the law of body weight and the law of 
body surface, both assume the simplest relationship between 
the size of the body and its heat-producing mechanism. Thus, in 
the law of body weight the natural assumption is made that 
each kgm. of body material has the same heat-producing power 
as every other kgm. of body material; in other words, that there 
is uniformity in heat production throughout the whole body. 
The second law, i.e., that the heat requirement of the body is 
directly proportional to its area, assumes that the basal heat loss 
of all individuals is constant per square meter of surface. Theses 
beliefs have so strong a hold upon modern writers that but few 
have given serious thought to the possibility of there being changes 
in the mass of active protoplasmic tissue in different bodies of 
the same size. The modern conceptions of the seat of com- 
bustion in the living body make almost unnecessary the assump- 
tion that heat production is greater in muscular tissue than in 
fatty tissue, and yet it is seemingly tardily admitted that inert 
body fat must l>e considered in a different category as to the 
function of heat production than either muscular or glandular 
tissue. Once this is admitted, however, the inadequacy of the 



28o Factors Affecting Basal Metabolism 

computation of the heat production per kfi,m. is obvious. For- 
merly the bones were also cbissified as inert material, but more 
recently McC'rudden^'* has maintained that the bones are fully 
as active in metabolism as are the muscles and the glands. 

Again it has been commonly considered in all the literature 
tlmt the active protoplasmic tissue or the heat-producing organ- 
ism of the body works with a constant intensity. In other words, 
little if any consideration is given to the possibility of there being 
various metabolic levels, or that the heat-producing organism 
per sc may functionate with a greater or less intensity. 

It is proper for us to consider here, therefore, first, what is the 
influence upon the basal metabolism of changes in the body com- 
position or changes in the relative proportion of active proto- 
plasmic tissue and inert body fat, and second, what are the stimuli 
to metabolism, noting if possible what variations in metabo- 
lism may be found in an organism which has essentially a con- 
stant mass of active protoplasmic tissue and a constant body 
surface, but with varying intensity of stimulus. 

Effect of variations in the 7fi<iss of active 'protoplasmic tisaiie. 

Variations in the proportions of the protoplasmic tissue and 
fatty tissue in the body may be expected with athletes, with 
men as compared with women, with men and women of similar 
weight but different height, and with certain infants. 

It is a well known fact that the process of athletic training 
removes inert body fat and increases and hardens the nmscular 
tissue, resulting in a greater proportion of protoplasmic tissue 
in the body. The study in the foregoing paper by Benedict 
and Smith^^ shows clearly that with trained athletes, particularly 
with the heavier men, the basal metabolism according to the 
three standards of comparison, namely, the total metabolism, 
the heat production per kgm. of body weight, and the heat pro- 
duction per square meter of body surface, is measurably greater 
with athletes than with non-athletic individuals of the same 
weight and height. Thus we have the first clear index as to the 

^* F. H. McCruddcn: TranHoctions of the ySth International Con'jrc^i^ 
on Ilyjiene and Demography, Washington, ii, p. 424, 1913. 
*• Benedict and Smith: loc. cit. 



F. G. Benedict 281 

definite influence of a greater proportion of muscular tissue upon 
the total metalx)lism. 

It is, furthennore, a well known fact that men as compared 
with women have a greater muscular development and are ca- 
pable of longer and more enduring muscular activity; consequently 
they do not, as a rule, have an excess of subcutaneous fat. In 
comparing a man and a woman of the same height and weight, 
therefore, one would normally expect to find a larger proportion 
of protoplasmic tissue with the man than with the woman. The* 
observations of Benedict and Emmes have shown clearly that 
when such comparisons are made, there is a distinct, although 
not necessarily striking, difference between the two sexes, the 
men having the gjeater metabolism. 

In the report of the study of the metabolism of normal and 
atrophic infants, Benedict and Talbot have pointed out that 
with two infants of the same height and weight, the older infant, 
who would naturally be atrophic, would frequently have the 
jjreater proportion of active protoplasmic tissue, while the younger 
would have the larger proportion of fat. It was found that th(» 
basal metabolism of the atrophic infant was always higher than 
that of the normal child. 

In comparing adults it is safe to state that with two individuals 
having the same weight but a difference in height, the taller 
individual will have the greater proportion of active protoplas- 
mic tissue, and the shorter the larger proportion of fat. We 
have therefore .compared in Table I a number of individuals of 
(iiflferent heights, but of the same weight. Anticipating the 
discussion in a succeeding section of the influence of age upon 
the metabolism, we have also selected for this comparison only 
those individuals who are approximately of the same age. Unfortu- 
nately the data for our eighty-nine male subjects, extensive though 
they are, do not permit a large number of comparisons of individ- 
uals having the same weight and age with widely varying heights; 
but such evidence as is given in Table I shows that the taller per- 
son has the larger heat production. 

In the comparison of normal infants with atrophic infants, of 
athletes with non-athletes, and of men with women, we have 
used individuals of like height and weight, and find a difference 
in the heat production of the two classes compared. In our 



282 Factors Affecting Basal Metabolism 



romparison of men of the same weight and age l)ut widely vary- 
ing heights, we have also found that the taller individual has 
a greater heat production than the shorter individual of the 
same weight. It is only reasonable to suppose that in all these 
comparisons we deal with actual diflferences in the total propor- 
tion of active protoplasmic tissue. The variations in the heat 
production may, however, be due, at least in part, to the fact that 
the active protoplasmic tissue functionates with a varying de- 
gree of intensity. 

TABLE I. 

Comparison of the heat production of normal men of like age and weight but 
of different height^ in experiments without food. 



SUDJECT 



F. G. B. 
Prof. C. 
K. H. A. 
F. P. R. 
W. G. J. 
S. A. R. 
R. I. C/. 
A. G. E. 

D. J. M. 

E. T. W. 



AQK 



41 
36 
26 
22 
21 
23 
26 
26 
20 
22 



NO. OF 
PERIODS 



37 

12 

no 

58 
26 
44 
9 
68 
31 
12 



HEIGHT 



cm. 



183 
169 
182 
173 
175 
165 
184 
169 
175 
169 



BODY WEIGHT 


TOTAL HEAT 
(COMPtrTED) 
PER 24 HR8. 


kgtn. 


cal. 


83.1 


1802 


83.0 


1655 


66.4 


1654 


65.1 


1543 


60.5 


1746 


60.8 


1460 


56.8 


1687 


57.0 


1531 


58.0 


1615 


57.8 


1472 



Effect of variations in stinudus. 

All comparisons of basal metabolism thus far made assume 
that the cellular activity of the protoplasm is constant in all 
cases, and the possibility of there being material differences in 
the metabolic level or the cellular activity under different con- 
ditions has not hitherto been given proper consideration. 

One of the first suggestions of a variation in the intensity of 
cellular activity is noted in Chart V, which shows the metabolism 
of our men subjects per square meter of body surface. In this 
chart we find that with eight men, weighing approximately 80 
kgm., the heat production per square meter of body surface 
ranged from 693 to 940 calories, and that five men, weighing 
50 kgm. or under, ranged from 693 to 958 calories. In other 



F. G. Benedict 283 

words, the two extremes of the heavy fat men and the thin small 
men show a wide range in the metabolism per square meter of 
lx)dy surface. This points strongly towards distinct differences 
in the intensity of cellular activity with different individuals 
rather than to corresponding mathematical differences in the mass 
of active protoplasmic tissue in the several groups, least of all 
to possible disturbances in the relationship between body weight 
and computed body surface. 

If we seriously consider the question of the possible stimuli 
to metabolism we find that it brings us immediately to a large 
number of possible variations in the metabolic intensity of nor- 
mal individuals. While the evidence presented in the preceding 
papers indicates differences in basal metabolism due to sex and 
to the muscular development of athletes, the popular conception 
that the vital activities of youth are greater than those of old 
age is certainly not without some foundation, and hence we 
should consider the influence of age upon the metabolism. 

Effect of age. 

Benedict and Talbot in their comparison of normal and atrophic 
infants have pointed out the influence of age upon the metabolism, 
but in comparing adults we have to consider not the effect pri- 
marily upon the proportion of muscle and fat, but the influence 
of age upon the body vigor and the cellular activity. In exam- 
ining critically all the foregoing charts, we find that those indi- 
viduals showing widest variations from the general course are 
somewhat frequently found among those over 32 years of age 
and those under 18 years of age. Furthermore the youths are 
grouped in a different part of the charts from the older men.^" 
While this was not invariably the case, yet we were sufficiently 
impressed with the fact to study more into the effect of age upon 
basal metabolism. Unfortunately the results of the obsei^va- 
tions made in this laboratory, while very numerous, do not con- 

'^ The number of plots on each chart is so great as to preclude indi- 
cating the initials of all individuals, but usually from the data in the 
charts and the large general table of Benedict, Emmes, Roth, and Smith 
(loc. cit.)f no difficulty will be experienced in recognizing the individual 
points. This is also true for the subjects of Palmer, Means, and Gamble. 



284 Factors Affecting Basal Metabolism 



3 



4 












CQ 
< 



ft = 



ft ^ 

ft 

o 






o 
* - s 

M 

S 






5 



£6 
a. M 



3 



V) 












s 



S 

2 
u 

s 

» S 



4; 









BS 






-: QC 
ft 

5 S 






00 
00 



5r 

o 

is 

a" 

< ^ 



I 



CO CO O M 



W 55 J5; 



tSSS 



CO o ic 



SIS* 



^ CO 00 
CO CO CO 



k-4 h-4 V^ 

l< ^i ^i 



00 > 

ao ^ 



3 



sg 



t- l^ t- 



O CI «p 

CD iQ IP 

c^ ei e5 



CO CO 



CI CI 

CO CO 



CI Q 

CO CO 



s^^^ & 



■>^ 35 r< £ 

w oe > 
X oc «>«• 






iC 



gff 



CI O Ci CO »C CO 



CO CO CO 00 Q CO 
CI CI CI Ci CO c« 



^ O "^ CS Q I>- 

~ '" ~ 1-M 00 CI 



O CO o 

CO I-" CI "^ 



CO 



CJ Q 1-* CI Ci "^ 

X O CO CO CO 



^ l>- CI QC O »C 

CO OS 00 O QO Q 

»o 'I* rp ""t^ f ^ 



CI CI CI •*!* f »C 
CI CI CI C> CI CI 



y ^ <>; e «: O 

;iL: ^ d ^ w 

ao a 00 00 CO K 

Z- 00 00 00 00 K 

• >rf •>i' (V^ i^M Oirf 'VM 

^H ^H irN ¥H •r* ^^ 

<*, .r^, *<, ,< <^. *< 



»0 CO I-" 
'-^ O 'f 
00 X oc 



oc OD 



^J4 



:3 



CI 



V 



X 



X 



-H Oi O 



l^ CO 



CO CI 



CI »0 ^ CO o o »« c <» 

b- 10 h- »A 00 CO X CO «0 

CI CI CI CI CI CI CI d M 

§ CI CI OS CO —< 1^ CO 

t^ CI "^ CS CO o i-- 

Tf »c »6 CO »c t>. »c 



OS OS CO CO CI C: iC ^ 
CO CO t-- t-- t^ CO !>• h- 



?5S 


s 


OS Ci 


* 


i-« CO 


M 


CO CO 


CO 


CI •- 




S I?? 




»— ^ «— 1 




CI --^ 




CO CO 





1^ 



y. 



M X 



l-^ CO 

OS CO 
10 »o 



CI X ^ 'r »ft o x; CO 

C5 r^ CO Q c t- !5 s 
•o »o lO ^ o «f? «* e 



CO CO 



a 
^ 



1 



CO t">» 



01 Tl Tl "^ CO CO '-' w 
CI CI 7\ CI CI CI CO ^ 



t^ t^ 






Si 

o 

v. 

e 

s 

^ . 



J- 

2 

a 

oc 

St 
V. 

< 



X 






'^s 






H W »-^ »-i W < -: -i 



Si. 
5 



S3 



> ^ 



V 



F. G. Benedict 



285 



9 



n S 00 $ 00 CO 8 
t« t^ r^ (» t^ CD b- 



i " 



S S 



00 Tt« iH 

• • • 

TT C^ to 

C9 CI el 






t» 


C4 CO 00 Tf o» 


s 


S 8 S 8 8 S 



«H t« kA ^^ io 

C4 c^ CD v5 o 

CO CO C^ CO o 



^ «D 
CO M 



CO »c 
t- 00 



s« s 



ss 



goo ^ CO cp ^ 
CD CD CO CO cD 






1-H 00 



1-t ^ kO <-H CO O) 



s 



C< Q 1-I Ol 00 
10 US kO kO ^ 



10 



So 






00 »A 10 t» 1^ "^ 
1-1 c5 ci c^ CO i>- 



s; 






-id 






> • 

00 
00 



o« 



iJ W -i 



1-* 00 CO iH 



i-j CO CO o 

t^ CO -^ to 

^1 CI C4 M 

IQ CI ^ 

9 S w 



q5 CO ^ 



00 

QQ 



s sss 



QQ 

QQ 



QQ 
QQ 



S & S 8 

t» t» CO t» 



CO 

00 



O t^ 00 

• • • 

O) O) Ok 

•-I 1-H ^H 

1/5 1—1 

o CO 




tC o S 



CO »o c« 



S3"^ 



00 



o ^ 



s 



CI 

00 



Si S 3 



CI 



^3 



. 07 






a, 
o 



C11 » 



0) 

> 



o 
08 *;? 

^ OQ 



I 

QQ 

•T3 



50 OQ 
08 -tA 

u o 

0) 0) 

>. — • 

< 



nnJOUBlf^LOr BIOLOOIC^LCBBMISTRT, VOL. XX, NO. 3 



286 Factors Affecting Basal Metabolism 

tain so large a number of people in middle age or old age or, 
indeed, in early youth as we would like. Nevertheless, there 
are a certain number of comparisons available and a few are 
given in Table II which permit the study of the influence of a 
difference in age upon the metabolism of individuals with the 
same body weight and height. 

It will be noted with the men subjects that in all cases the 
youths had a very much higher metabolism than the adults 
with whom they were compared. Thus, in the first group the 
two boys showed a considerably higher metabolism than that 
of the adults in the same group. In Group II both boys showed 
a very much higher metabolism than the other two subjects, 
while in Group III the heavy young man, O. F. M., 24 years old, 
showed a perceptibly higher metabolism than the two older men 
36 and 63 years, respectively, with whom he was compared. The 
average of all the younger subjects both per kgm. of body weight 
and per square meter of body surface was considerably higher 
than that of the older subjects. 

With the women subjects the difference was not so sharply de- 
fined. In Group I the 15 year old girl had a much higher metab- 
olism than the other subjects; in Group II the metabolism of 
the girls was essentially the same as that of the women. In 
Group III the 18 year old girl had a metabolism slightly lower 
than the average of the women in the group with whom she 
was compared, but in Group IV the 24 year old woman had a 
much more active metabolism than the 52 year old woman with 
whom she was compared. Here again, the general average, de- 
fective though it obviously is, shows a distinctly greater metab- 
olism for the younger subjects. 

This difference in metabolism of individuals of different ages 
was early noted by Sond^n and Tigerstedt,^^ who conclude that 
age and especially the age of growth has a great influence upon 
the total metabolism of the body, and that the younger organism 
has the greater metabolism per unit of body surface than has 
the older. 

Similarly Magnus-Levy and Falk" noted that the metabolism 

" K. Sond^n and R. Tigerstedt: Skand. Arch,f, PhynoL, vi,- p. 218, 1805. 
"A. Magnus-Levy and E. Falk: Arch, f, AnaL u. Physiol, , Physiol, 
AbUf Supplement, p. 314, 1899. 



F. G. Benedict 287 

of children is greater than that of adults and that the metabolism 
decreases in old age. This finding likewise applied to the meas- 
urements on the basis of heat per square meter of body surface. 

It is somewhat difficult to differentiate sharply between the influ- 
ence of variations in the total proportion of active protoplasmic tis- 
sue and the influence of variations in the stimulus to the cellular 
activity, and to say, for example, whether in old age the lower 
metabolism is due to a lower cellular activity or to an actual 
decrease in active protoplasmic tissue from disintegration and 
loss and atrophy of the tissues. A fact of fundamental impor- 
tance, however, is that there may be, certainly in the cycle of a 
lifetime, very considerable changes in the metabolism of an 
individual. The youth has a higher metabolism than the per- 
son in middle life, while one of advanced years has a still lower 
metabolism than the person in middle life. Thus far the history 
of investigations in metabolism has been written in too short 
a time to enable an accumulation of accurate scientific data 
concerning the metabolism of the same individual during va- 
rious periods of life. Practically no investigations include a 
study of the metabolism of a subject in early youth as compared 
with his metabolism in middle life or later. 

It is not necessary, however, to await the entire period of a 
lifetime to note differences in the intensity of cellular activity; 
for there are numerous factors which may produce in a short 
time changes in metabolism comparable to those noted with 
people of different ages. Among these may be mentioned sleep, 
character of diet, and after-effects of severe muscular work. 

Influence of sleep. 

It is possible to maintain complete muscular repose and yet 
have the brain active and awake; on the other hand, we may have 
complete muscular repose with the subject sleeping. Observa- 
tions on deep sleep made in earlier investigations have led to the 
almost unanimous belief that sleep has no influence upon the 
metabolism. 

A study of prolonged fasting recently carried out in this lab- 
oratory afforded an excellent opportunity for comparing the metab- 
olism during the night when the subject was sleeping quietly 



288 Factors Affecting Basal Metabolism 

in the bed calorimeter with that of the next morning when he 
was lying quietly upon the same bed, awake and breathing into 
the universal respiration apparatus. The subject slept for the 
greater part of the period of observation in the bed calorimeter, 
the graphic record of the body movements made by the self- 
recording bed showing that the man was remarkably quiet 
throughout the whole night. During the morning observation, 
when the subject was connected with the respiration apparatus, 
he was phenomenally quiet, the graphic record showing a prac- 
tically straight line in every experiment. According to the 
opinion of Mr. T. M. Carpenter, who made the observations 
with the respiration apparatus, the subject had the most com- 
plete muscular relaxation and control of any of the individuals 
that he had ever studied. The details of the observations with 
this subject may be found in the report of this fasting experi- 
ment,^ but it may be stated here that during the thirty-onti 
days of fasting the metabolism gradually decreased. Without 
taking account of the changes in the body weight, we may com- 
pare for each day the metabolism of the subject while in the beil 
calorimeter during the night with his metabolism immediately 
afterward when he was connected with the respiration apparatus 
in the morning. These comparisons, which are given in Table 
III on the basis of the oxygen consumed, show in general an 
increase of 13.2 per cent for the morning metabolism when tlu» 
subject was connected with the respiration apparatus. 

The increased metabolism during the morning observations 
cannot be attributed to muscular activity; for a comparison of 
the graphic records shows that the degree of muscular repose* 
was more nearly perfect in the morning experiments with th(» 
respiration apparatus than in the night experiments with the 
bed calorimeter, since it was naturally impossible for the sul>- 
ject to lie absolutely quiet throughout the whole night, even dur- 
ing sleep. 

There is no question of the influence of food in the alimentary 
tract; for during the entire period of thirty-one days the sub- 
ject ate absolutely no food and drank only about 900 cc. of dis- 
tilled water per day. 

»' F. G. Benedict: Carnegie Institution of Washington Publications, 
No. 203, 1915. 



F. G. Benedict 



289 



Inereast in nuUAolitm 0/ lubjeet aviake as compared vdtk metaboliim of 
lubjeel CM jeep. 





(.) 




SubjettMleep 




«. 


iBt 


196 


3d 


208 


3d 


198 


4th 


187 


5th 


176 


6th 


185 


7th 


185 


8th 


177 


Qth 


173 


10th 


179 


tlth 


166 


I2th 


173 


13th 


167 



Uth 
15th 
16th 
17th 
tSth 
19th 
30th 
3Ut 



14.1 
13.4 
16,5 



10,5 
15 2 
18.2 
13.0 

15.7 



34th 
25th 
26th 
27th 
28th 
29th 
30th 

3lBt 



12,9 
11,3 
IS 6 

14,5 



290 Factors Affecting Basal Metabolism 

Since, therefore, the subject was asleep for the most pai't of 
the caloruneter observations in the night period and awake 
during the observations with the respiration apparatus in the 
morning period, and the metabolism was not influenced by mus- 
cular activity or the ingestion of food, it is logical to conclude 
that the increased metabolism during the morning observations 
was due to the fact that the subject was awake. The experi- 
mental data therefore justify the conclusion that deep sleep 
lowers perceptibly and very considerably the basal metabolism, 
and we may properly ask if our standard for the measurement 
of the basal metabohsm is the correct one. We state as the 
onlj' prerequisites of the measurement a post-absorptive con- 
dition and complete muscular repose, thereby eliminating the 
influence of the ingestion of food and the influence of external 
muscular work. Is it not conceivable that we should logically 
eliminate the question of increase in the internal muscular activity 
incidental to the waking condition, and consider the basal metal>- 
olism to be that obtaining during the post-absorptive condi- 
tion, with complete muscular repose, and in deep sleep? 

Variations in metabolism as observed at different times. 

While it has thus far been impracticable to make studies of 
the same individual during early youth, middle age, and old 
age, it is important for us to note the significant changes in metal>- 
olism which may be observed with the same individual on con- 
secutive days. As a result of observations by Zuntz and some of 
his students and the earlier published reports of Benedict and 
Carpenter, it has been commonly considered that the metabolism 
of an individual remains essentially constant from day to day. 
In the accumulation of the experimental data for this study, 
opportunities occurred to note the actual variations in the metab- 
olism of individuals under the predetermined conditions of the 
post-absorptive condition and complete muscular repose. With 
many of these subjects observations were made on five or more 
days and frequently over periods of several months and even 
years. The results for a large numl)er of our subjects who have 
been studied five daj-s or more have l)een gathered top:ether in 
Table IV. 



F. G. Benedict 



291 



TABLE IV. 
Variation in post-absorptive metabolism in experiments with normal men. 





1 


OB8EBVATIONB 


t 

1 

Tim 


VABIATION IN 


SUBJECT 


▲GB 


Days 




COTXBXD BT 
■XPBBXMBMTB 


OXYGEN ABOVE 




Periods 


lOKIlllTM 




yrs. 






percent 


F. G. B. 


1 ^^ 


8 


37 


2 yrs.2mos. 


10.8 


E. G. 


! 20 


6 


11 


lyr. 


11.2 


C B. S. 


26 


26 


75 


1 yr. 2mo8. 


18.2 


J. 11. xl. 


25 


5 


13 


4mos. 


j 14.1 


K. H. A. 


26 


25 


110 

1 


llmos. 


19.3 


J. R. 


! 27 


12 


57 


3mos. 


21.2 , 


M. A. M. 


1 29 


53 


157 


4mos. 


21.3 


F. P. R, 


! 22 


20 


58 


3mos. 


15.2 


J. J. (^. 


27 


' 53 


252 


2 yrs. 3iD08. 


12.1 


D. M. 


22 


! 5 


15 

1 


5dys. 


15.3 


M. J. S. 


24 


13 


42 


19 dys. 


13.6 


M. Y. B. 


20 


6 ! 


12 


6dy8. 


8.1 


VV. F. M. 


21 


7 


12 

1 


15 dys. 


16.7 


H. H. A. 


22 


i 28 


81 ! 


1 yr. 2mo8. 


24.8 


S. A. R. 


23 


' 9 


. 44 


5mos. 


16.6 


W. G. J. 


21 


9 


26 , 


Imo. 


10.6 


H. L. M. 


26 


35 


120 


2 yr8.5mos. 


25.9 


J. E. F. 


21 


7 


24 


5mos. 


22.7 


J. K. M. 


24 


27 


103 : 


1 yr. 2mos. 


8.2 


J. B. T. 


20 


11 


^ 


8mos. 


8.9 


W. F. B. 


32 


5 


16 1 


4mos. 


9.3 


M. B. L. 


20 


5 


18 


2mo8. 


8.6 


Li. E. E. 


31 


31 


144 > 

1 


2 yrs. Imo. 


19.8 


Dr. S. 


43 


5 


13 i 


12 dys. 


4.9 


D. J. M. 


20 


5 


31 


2mos. 


7.7 


H. F. T. 


32 


41 


211 


8mos. 


31.3 


P. F. J. 


20 


18 


82 


9mos. 


10.0 


A. G. E. 


26 


14 


68 


lyr. 


7.7 


W. A. S. 


21 


5 


10 


2mos. 


5.4 


Dr. P. R. 


41 


9 


33 


7mos. 


11.6 


C H. M. 


19 


9 1 


25 


4mo6.' 


12.1 


V. G. 


17 


17 


71 i 


5mos.- 


15.0 


J. J. G. 


21 


6 


20 


2mos.i 


9.6 


T. M. C. 


35 


17 


93 


4 yrs.5moB. 


14.9 


J. H. 


26 


6 


12 


12 dys. 


3.5 


Average 


- — • 


1 




; 




(35 sub- 












jects) 




I 






13.9 



292 Factors Affecting Basal Metabolism 

The metabolism in all cases was most carefully determined^ 
so that variations in the results were not due to errors in tech- 
nique, but to variations in the basal metabolism under the experi- 
mental conditions outlined. In the experiments with the res- 
piration apparatus a tendency to fall asleep has been shown by 
many subjects, the sound of the motor and blower producing 
a soothing effect; consequently it is not possible to stat« that 
in every experimental period there was complete consciousness 
on the part of the subject. In a large majority of the experi- 
mcnts, however, the subject was fully awake and the variations 
noted between the maximum and minimum metabolism can- 
not fairly be stated as due to sleep. 

As an index of the variations in metabolism during the time 
covered by the experiments, we have taken the increase in the 
oxygen consumption, using the mmimum value as a basis. With 
these subjects it will be seen that in one case the oxygen con- 
sumption varied as widely as 31.3 per cent, while with still another 
subject it varied only 3.5 per cent. A general inspection of the 
data will show that, as a rule, the greatest variations were found 
with the subjects studied over the longest periods. While it is 
hardly correct to obtain an average value for the oxygen con- 
sumption for so many different individuals with such wide differ- 
ences in the time covered by the experiments, yet such a value 
has been found and shows that on the basis of these observations 
there may be an average variation of 13.9 per cent in the basal 
metabolism, when measured in the post-absorptive condition 
and with complete muscular repose, during a period of two years 
or, in the majority of cases, considerably less. With no attempt 
to analyze the causes of these differences, it is sufficient here 
simply to call att<5ntion to their magnitude. 

Examination of Table IV shows clearly the error of assuming, 
as Is frequently done in metabolism experiments, a basal value 
for any one individual to compare with metabolism measure- 
ments made with a superimposed factor. Since this is illogical 
in the case of one individual, it is even more illogical to con- 
sider possibilities of a standard normal value for any group of 
individuals. 

It is relatively rare for the metabolism of an individual to be 
determined for the same periods on consecutive days, and hence 



F. G. Benedict 293 

it is difficult to state whether the stimulus to the cellular activity 
fluctuates very considerably from day to day or whether the 
variation in the metabolism is due to periodicity, weather, tem- 
l^erature conditions and changes, or similar factors. One of 
the most extensive series of observations in which the basal 
metabolism of an individual was determined practically every 
day is that reported by Benedict and Cathcart^^ with a profes- 
sional athlete, M. A. M. With this subject the basal metabolism 
was determined almost daily for the period from December 7, 
1911, to February 29, 1912, with a few observations after that 
date. Certain of the experiments were complicated by the 
fact that on the preceding day there were considerable variations 
in the character and nature of the diet, as the subject was given 
on some days a diet with but 100 grams of carbohydrate and 
on other days a diet with 400 grams. Excluding these days 
with special diet, we find the body weight varied from 64.5 kgm. 
to as high as 67.2 kgm. Excluding, also, the first day of experi- 
mentation, on which the conditions were admittedly unusual, 
we find during this period a minimum carbon dioxide produc- 
tion of 191 cc. per minute and a maximum production of 232 
cc. per minute. The oxygen consumption during the same 
period ranged from a minimum of 225 cc. per minute to a maxi- 
mum of 262 cc. per minute. The average values were 206 cc. 
per minute for the total carbon dioxide production, and 240 cc. 
per minute for the total oxygen consumption. 

In this connection we may also refer to the recent article of 
Palmer, Means, and Gamble, which gives the results of a series 
of observations on W. W. P. for six days, from July 21 to July 
26, 1914. The agreement of the values is as close as one could 
expect, and taking into consideration these values alone, one 
could readily assume a constant metabolic activity from day to 
day. On the other hand, if we examine the values obtained with 
W. W. P. in the winter of 1913-1914, which are given in the 
first general summary table of the article by Pahner, Means, 
and Gamble (their Table I), we find marked differences from 
the values secured in the summer of 1914. In the winter the 
body weight was essentially the same as in the summer, njimcly, 

" F. G. Benedict and K. P. Cat heart: ibid.. No. 187, p. 78, 1913. 



294 Factors Affecting Basal Metabolism 

93.9 kgm. as compared with an average of 93.7 kgin.; but in 
the winter the total heat production was 2004 calories and in 
the summer 1797 calories; the heat production per kgm. of body 
weight was 21.4 calories in the winter and 19.2 calories in the 
summer; and the heat production per square meter of body 
surface was 789 calories in the winter and 707 calories in the 
summer. Thus in this most recent research we have indications 
of a marked difference in the basal metabolism as determined 
in the post-absorptive condition and with complete muscular 
rest. It should be stated, furthermore, that the series of experi- 
ments made in the winter of 1913-1914 was not the first made 
with W. W. P. with this apparatus, and the results therefore 
represented the metabolism of a more or less trained subject. 
Obviously he did not change in height in the period interven- 
ing between the two series, and, as has been shown, the weight 
did not change. His habits of life were such that probably 
there was no replacement of active protoplasmic tissue by inert 
fat. We are consequently forced to the conclusion that wo 
have here not alterations in the amount or proportion of thii 
protoplasmic tissue, but a distinct variation in the stimulus 
to cellular activity. 

It should not be lost sight of that the fact that W. W. P. had 
a higher metabolism in the winter of 1913-1914 than in the 
summer of 1914 may be taken by some investigators as an indi- 
cation of a larger metabolism due to an increased cooling of 
the body surface, and that we have here the possibility of a 
greater temperature difference with a greater metabolism. As 
a matter of fact, the temperature of the laboratory during both 
series of experiments was essentially constant, the diflferen(*e 
being but a relatively few degrees. Furthermore, it is well 
known that the human body does not react to differences in tem- 
perature environment as do the lower animals. Thus the evi- 
dence of both Loewy^ and Johansson-® is strikingly against there 
being any material alteration in the metabolism of man with 
cold until the temperature difference is sufficiently great to 
induce internal muscular work due to shivering. 

"A. Loewy: Arch. f. d. ges. Physiol. ^ xlvi, p. 189, 1890. 
'* J. E. Johansson: Skand. Arch. /. Physiol. f vii, p. 123, 1897; xvi, p. 
88, 1904. 



F. G. Benedict 295 

Preliminary observations have been made upon dogs in this 
laboratory to note exactly the influence of temperature environ- 
ment upon the metabolism. When the animals are suspended 
in a crib or cage, by means of which a graphic record can be 
obtained of the muscular activity, it has been found that with 
no muscular movement or shivering, very much greater differ- 
ences in temperature environment may be borne without change 
in the metabolism than had hitherto been supposed. In at 
least one instance a dog in this laboratory has shown no altera- 
tion in the metabolism with a difference in the external environ- 
ment of 10°C. Only with the onset of shivering does the metab- 
olism increase perceptibly. It should be stated, however, that 
this investigation is by no means complete, and it is not the 
purpose here to enter into a discussion of the influence of tem- 
perature environment upon the heat production of lower ani- 
mals. With men the evidence points strongly towards the con- 
stancy of metabolism irrespective of moderate changes in at- 
mospheric environment. This has likewise been borne jout by 
the results of experiments made by Schlossmann and Mursch- 
hauser^^ on sleeping infants. 

Diurnal variations in meiaholism. 

The routine for conducting experiments for the determination 
of the basal metabolism usually involves a series of experiments 
in the early morning in the post-absorptive condition; that is, 
before food is taken. Very little data are available to show 
whether or not the course of the metabolism throughout the day 
is altered materially. The best evidence that the Nutrition 
Laboratory possesses is that obtained with the fasting subject, 
who, as has already been noted, had a considerably greater metab- 
olism awake than asleep. Experiments made with this sub- 
ject in the afternoon after a day spent in talking and various 
experimental tests, but with little muscular exercise, showed that 
he invariably had a higher basal metabolism than he did in the 
forenoon. While of course the metabolism of a subject living 
under these artificial conditions may not be compared with that 

*^ A. Schlossmann and H. Murschhauser : Biochem. Ztachr., xxxvii, 
p. 1, 1911. 



296 Factors Affecting Basal Metabolism 

of a normal individual, nevertheless it is a fact that this organ- 
ism which, aside from the absence of food, was othen^'ise normal, 
had three sharply defined metabolic planes. These values were: 
first, a value when the subject was asleep in the bed calorimeter; 
second, a value obtained when he was awake lying on a mattress 
about 9 o'clock in the morning; and third, a value obtained 
under the same conditions in the late afternoon. Using as a 
basis the metabolism for the night during sleep, we find that 
the metabolism in the morning with the subject awake had in- 
creased 14 per cent, and in the late afternoon, under the same 
conditions, had increased 22 per cent. It is furthermore of 
interest to note that many of the infants studied by Benedict 
and Talbot exhibited diflferences in metabolic levels, as indi- 
cated by an increased pulse rate and an increased gaseous metab- 
olism, even though in sound sleep and with no evidence on the 
kymograph record of muscular activity. 

• Sthmdu^s as influenced by prolonged fasting. 

Further striking evidence of the probable effect of a decreased 
stunulus is found with the fasting subject, who showed a de- 
pression in the metabolism wholly out of proportion with the 
changes in the body weight or in the body surface as computed 
l)y the Meeh formula. The heat production per square meter 
of body surface as computed on the morning of the first fasting 
(lay, that is, about eighteen hours after the last food, was 859 
calories. As the fast progressed, the heat production on this 
basis of computation fell until a minimum of 668 calories was 
observed on the morning of the twenty-third day of fasting. 
No disproportion between the body weight and body surface 
could bo assumed with this individual corresponding to this 
difference in heat production, and we are again convinced of the 
fact that here we deal with a variation in intensity of a true 
stimulus to cellular activity. This is furthermore emphasized 
by the fact that after the twenty-third day there was a distinct 
tendency for the metabolism to rise which was accompanied by 
a measurable increase in the pulse rate. Thus the tendency 
to depress the metabohsm, due to the continued loss of proto- 
plasm by fasting, was actually overcome by the unknown stimu- 



F. G. Benedict 297 

lus increasing the cellular activity of the remaining body sub- 
stance, ultimately resulting in a positive increase in the basal 
metabolism. 

Influence of character of diet. 

While it may seem questionable to introduce here a discussion 
of the influence of a carbohydrate-free diet upon the post-ab- 
sorptive metabolism, it has been demonstrated that with the acid- 
osis resulting from the ingestion of a carbohydrate-free diet, there 
is a distinct increase in the basal metabolism of normal indi- 
viduals. This increase is clearly not due to a change in the 
body weight or the body surface or to variations in the mass of 
active protoplasmic tissue, but must be due to an alteration 
in the stimulus to cellular activity, the presence of acids in the 
body stimulating the cellular activity to a higher level. 

After-effects of severe muscvlar work. 

Finally, mention should be made of the marked after-effects* 
of severe muscular work noted by Benedict and Cathcart*^ with 
their professional athlete, M. A. M., who had not partaken 
of food for nearly twenty hours, during which time he had 
performed an enormous amount of muscular work. After the 
cessation of work, the metabolism showeH a prolonged though 
steadily decreasing influence of the preceding muscular activity. 
The investigators believe that this stimulus to the cellular activ- 
ity continued for a long time after all external evidence of mus- 
cular activity had ceased. Irideed, so long did the stimulus 
continue that some writers might ascribe at least a part of the 
increased metabolism noted with athletes as compared with 
normal individuals to the possibility of the after-effects of the 
muscular activity on the preceding day. Of most importance, 
however, is the fact that as a result of excessive muscular activity 
the cellular activity may be increased enormously and main- 
tained at a high level for a considerable period after the cessa- 
tion of the work, thus clearly establishing a higher, though con- 
tinually decreasing, metabolic plane. 

*• Benedict and Cathcart: loc. cit. 



298 Factors Affecting Basal Metabolism 

SUMMARY. 

It is thus obvious that in studying the basal metabolism of 
a normal individual, we have a much larger number of factors 
to deal with than has hitherto been recognized. 

Unquestionably body weight plays an important part. In 
general, large bodies give off larger amounts of heat than smaller 
ones, but there is no direct relationship between the total body 
weight and the total heat production. 

« 

. The mathematical relationship between the body surface and 
the body weight established by direct measurement has led to 
the general belief that the heat production is determined by the 
body surface^ since an approximate relationship has been fre- 
quently noted between them. Careful analysis of metaboUsm 
measurements obtained on athletes, normal men and women, 
and normal and atrophic infants, leads to the conclusion that 
the metabolism or heat output of the human body, even at rest, 
. does not depend upon Newton's law of cooling, and is, there- 
fore, not proportional to the body surface. 'While certain dis- 
turbances in this supposed relationship between the heat pro- 
duction and the body surface may correctly be ascribed to errors 
in the formulae used for computing body surface, nevertheless 
the vast bulk of the. evidence, particularly with athletes and 
with infantSj and to a considerable extent with so called normal 
individuals, shows that the variations between metabolism and 
body surface are far outside of any possible errors in formulae. 
Body composition, i,e;.j the proportion of inert body fat and 
active protoplasmic tissue, has a great influence upon the basal 
metabolism. The tendency toward the greater metabolism 
shown by athletes in comparison with non-athletes may thus 
be explained by the greater muscular development as indicating 
a larger proportion of active protoplasmic tissue. The apparent 
influence of sex, as brought out in the comparison of the metab- 
olism of men and women, may also be attributed to the greater 
proportion of inert body fat in the latter, with a consequent 
smaller aQiount of active protoplasmic tissue. It has also been 
.seen that (height is a factor in determining the basal metabolism, 
since in comparing individuals of like body weight and age, but 
widely varying height, the taller individual has usually the 



F. G. Benedict 299 

greater metabolism; this is likewise due without doubt to the 
fact that the taller individual has the larger amount of active 
protoplasmic tissue. All these variables deal directly with the 
mass of the heat-producing organism; ?.e., the amount of active 
protoplasmic tissue. 

We have still another very important factor; immely, the 
stimulus to cellular activity. This stimulus is influenced by a 
number of factors. One of these factors is age, and it has been 
noted that with the growing organism of youth, there is a much 
greater cellular activity than with the adult, and a consequent 
higher metabolism. It has been brought out, however, that 
in old age there may be actual atrophy of protoplasmic material. 

Sleep has also been shown to have an influence upon the basal 
metabolism, the stinaulus to the cellular activity being greater 
with an individual when he is lying awake than when he is asleep. 

Considerable fluctuations in the basal metabolism havB been 
found from day to day not only with a fasting man but with 
normal individuals studied over considerable periods of time. 
These variations could not logically be attributed to changes 
in lx)dy weight or body surface, and obviously there was no change 
in height. Even in the course of twenty-four hours, the fast- 
ing subject was found to have three distinct metaboUc planes, 
showing clearly a diurnal variation in the stimulus to the cellu- 
lar activity. 

Still other factors considered as influencing the stimulus to 
cellular activity are prolonged fasting, the character of the pre- 
ceding diet, and the after-effects of severe muscular work. 

From the evidence gathered with the various subjects studied, 
it is clear that the basal metabolism of an individual is a func- 
tion, first, of the total mass of active protoplasmic tissue, and 
second, of the stimulus to cellular activity existing at the time 
the measurement of the metabolism is made. Apparently at 
present no law can be laid down that will cover both ©f these 
important variables in the basal metabolism of an individual. 



A RESPIRATION APPARATUS FOR SMALL ANIMALS. 

By FRANCIS G. BENEDICT. 

(From the Nutrition Laboratory of the Carnegie InstitiUion of Washington, 

Boston.) 

(Received for publication, January 15, 1915.) 

The increasing use of determinations of the respiratory exchange 
as an index of alterations both in the character and in the amount 
of the basal metabolism calls for an apparatus that may be 
satisfactorily used for small animals, with which the carbon 
dioxide output and the oxygen intake may be determined with 
sufficient accuracy to give reliable respiratory quotients. The 
determination of the carbon dioxide output is relatively simple 
with small animals, and a number of methods have been used 
to secure this end. Perhaps the simplest is the closed chamber 
respiration apparatus, in which the carb6n dioxide is allowed to 
accumulate and samples of air are analyzed from time to time. 
Another method conmionly used is to pass a current of outdoor 
air through the respiration chamber, measure the amount of air 
passing through, and note the increment in the carbon dioxide 
percentage. 

On the other hand, the determination of the oxygen consumption 
of small animals presents technical difficulties so great that it is 
rarely attempted. In a relatively few experiments a tracheal 
cannula has been used for short periods of observation, the 
inspired air being separated from the expired air by a system of 
valves; the expired air may thus be measured, sampled, and ana- 
lyzed. This method has recently been employed by Tangl,^ but 
nearly all his animals weighed 5 kgm. or over. The indirect 
method of Haldane,^ and more recently the combined direct and 
indirect method of Fredericia' have also been occasionally em- 
ployed for the determination of the carbon dioxide output and the 

* F. Tangl: Riochcm. Zlschr., xxxiv, p. 1, 1911. 

' J. S. Haldanc: Jour. Physiol. y xiii, p. 419, 1892. 

* L. S. Fredericia: Biochem. Ztschr., liv, p. 92, 1913. 

301 

THB JOURNAL OF BIOLOGICAL CHCMISTBT, VOL. XX, NO. 3 



302 Respiration Apparatus for Small Animals 

oxygen consumption of small animals; but the determination of 
these factors for animals as small as a rabbit or a guinea pig has 
not been common in laboratory technique. 

Ebcperiments with small animals are of value in the investi- 
gation of many problems and may be extended twenty-four 
hours or more, thus permitting the collection of appreciable 
amounts of carbon dioxide and the absorption of corresponding 
amounts of oxygen. Such long periods of observation will neces- 
sarily be accompanied by more or less muscular activity. It is 
becoming more and more evident to workers in metabolism that 
in studying basal values or the superimposition of a factor affect- 
ing metabolism upon the basal value, only periods of complete 
muscular repose may advantageously be used; on the other hand, 
there are numerous problems, particularly those in which the 
character of the katabolism is to be studied, which do not require 
absolute muscular repose for their investigation, although it is 
always desirable. For such experiments a respiration apparatus 
or, more properly, a modification of the universal respiration 
apparatus, has l)een used in this laboratory for the measurement 
of both the carbon dioxide production and the oxygen consump- 
tion of animals as small as a rabbit or a guinea pig. A description 
of this modified apparatus follows. 

The universal respiration apparatus has been extensively em- 
ployed in this laboratory for the past few years. It has been 
readily adapted to experiments with a man at work and has been 
successfully employed for such experiments when the carbon 
dioxide production was as great as 2500 cc. in one minute. It has 
also been applied to the measurement of the metabolism of in- 
fants. The universal respiration apparatus is now established 
in several other physiological laboratories. Thus far it has been 
used in these laboratories primarily for studies on men, and when 
such studies are not in progress, it has been unemployed. I 
purpose showing here how the apparatus may be so modified by 
a simple valve system that it may also be readily applied to the 
measurement of the metabolism of animals weighing but 1 or 
2 kgm. 

With the univei*sal respiration apparatus,* as used most fre- 
cjucntly for studying the respiratory exchange of man, the subject 

* F. G. Benedict: Deutsch, Arch. /. klin. Med., cvii, p. 166, 1912. 



F. G. Benedict 



303 



breathes through a nosepiece or a mouthpiece and draws air 
into the lungs from the ventilating air current. The displaced 
oil ultimately passes through a carefully counterpoised spirom- 
eter which moves up and down with each respiration. By 
connecting a small respiration chamber to the ventilating air 
circuit and allowing the spirometer to act as a supplementary 
expansion chamber, air may be circulated through the respiration 
chamber at any desired rate of speed. An animal confined in the 
respiration chamber gives off carbon dioxide with each expi- 




Fia. 1. KJchematiu 



of respiration appar.itus for small aniraala. 



ration, which is later removed from the ventilating air current; 
it also absorbs oxygen with each inspiration, this consumption 
of oxygen being indicated by the decreased volume of air in the 
spirometer; the loss of oxj-gen from the air is ultimately measured 
by noting the amount required to supply the deficiency. A 
diagram of the apparatus as used for experiments with small 
animals is given in Figure 1. 

General description of apparatus. The course of the circulating 
air current is shown in the schematic outline of the absorbing 



304 Respiration Apparatus for Small Animals 

apparatus at the right side of the figure. This does not differ 
in general from that illustrated in previous communications in 
which this form of apparatus has been described as used with man 
and with infants. The ventilating current is kept in motion by 
a rotary blower B, which is driven by an electric motor. The air 
leaves the chamber at the point A and, after passing through 
the blower, is forced first through an empty glass vessel C, which 
serves as a safety trap, and then through the two so called ** Wil- 
liams bottles" D and E containing sulphuric acid which removes 
the moisture. The air, which is now water-free but contains the 
carbon dioxide produced by the animal and also lacks the oxygen 
consumed, passes into the 2- way valve Vu where it may be de- 
flected into the bottle F containing soda lime for the absorption 
of the carbon dioxide. Since the soda lime is necessarily some- 
what moist, moisture is yielded to the air in its passage through 
the vessel Fy which must be collected or otherwise it causes an 
error in the measurement of the carbon dioxide. This moisture 
is absorbed by the sulphuric acid in the accompanying bottle G. 
The air then passes through a second 2-way valve F2 to a small 
brass can H, containing dry sodium bicarlx)nate, this salt remov- 
ing the unweighable traces of acid vapor retained by the air as 
it passed through the vessel G. Sufficient moisture for comfort- 
able breathing is supplied to the air in its passage through the 
water in the glass vessel J; the proportion of the air passing 
through the water in J is controlled by means of the by-pass 
valve K, The air is then returned to the respiration chamber. 
This chamber, which is cylindrical in form, is constructed of copper 
and has a cover M with a water seal. The glass plate A^ in the 
cover gives opportunity for observation of the animal. 

Spirmneier, As thus arranged the entire ventilating system 
and the chamber are of rigid wall construction and permit no 
automatic fluctuations of the volume such as would be caused 
by changes in barometric pressure or temperature. The spirom- 
eter of the universal respiration apparatus, which provides for 
such fluctuations in the volume of the air, is usually attached to 
the ventilating system near the pipe on the intake side of the 
blower, and in Figure 1 it is shown as attached to a 3-way valve 
at this point (0). Ordinarily the spirometer is so connected that 
the entire volume of the ventilating air current passes through the 



F. G. Benedict 305 

spirometer, but in the adaptation of the apparatus for experi- 
ments with small animals, this is unnecessary. As the volume 
of air in the chamber changes by reason of changes in the baro- 
meter, in the temperature, or on account of the absorption of 
oxygen by the subject, the spirometer freely adapts itself to the 
variations in volume, so that the pressure inside the chamber is 
alwajrs the same as that of the atmosphere. From the readings 
of the height of the spirometer on an attached scale, a mathe- 
matical estimate of the variations in the volume may be readily 
obtained. Oxygen to replenish that used by the animal is added 
through the pipe P whenever such need is indicated by the height 
of the spirometer bell. 

Thus the respiratory system of this apparatus consists of a 
chamber,* ventilating air current, abisorbing vessels to remove 
the carbon dioxide and water, spirometer to provide for changes 
in the volume and for indicating the need of oxygen in the chamber, 
and a tube for the introduction of oxygen. 

Rate of ventilation. The Crowell rotary blower, commonly 
supplied for the univeraal respiration apparatus, is capable of 
maintaining a ventilating air current of 100 liters per minute, 
such as would be used in experiments with severe work. In 
experiments with smair animals, such a high rate of ventilation 
is obviously unnecessary. Indeed, the rate employed for respi- 
ration experiments with a resting man, i.f ., 35 liters per minute, is 
much too large. It has been found impracticable to reduce the 
speed of the blower so that it would deliver but 2 to 3 liters of 
air per minute, but the effective ventilation of the chamber may 
be reduced to any desired amount by means of the by-pass valve 
Q. When this valve is open, the air circulates only around the 
blower and is not forced through the ventilating system. By 
partially closing the valve any desired rate of ventilation may be 

^ All our observations on single rabbits and guinea pigs have been made 
with a chamber having a volume of 17 liters. Obviously the smallest 
volume consistent with the size of the animal is to bo desired. By placing 
the water seal at the bottom and varying the size and shape of the covers, 
various heights and sizes of chambers may be constructed and readily 
attached to the absorbing apparatus. All pipes may advantageously re- 
main in the bottom of the chamber. Internal supports of different heights 
for the various sizes of swinging cages present no difficulties to the mechani- 
cian. 



3o6 Respiration Apparatus for Small Animals 

readily obtained. The ordinary rate required for a rabbit weigh- 
ing 1 or 2 kgm. would be approximately 2 liters per minute. 

Measurement of the carbon dioxide produced. The amount of 
carbon dioxide produced is determined by weighing the soda 
lime bottle F and its accompanying sulphuric acid bottle G on a 
balance which is sensitive to 0.01 of a gram, the Sauter balance 
being found the most satisfactory for this purpose. At the end 
of any given experimental period, the air current may be de- 
flected to a second set of carbon dioxide absorbers by turning 
the 2-way valves Vi and Vt simultaneously, and the experiment 
proceeds without mterruption. 

Measurement of the oxygen consumption. The amount of the 
oxygen consumed may be determined in a nimiber of wa3rs. The 
spirometer of the universal respiration apparatus has a flexible 
volume of not far from 2 to 3 liters. This spirometer can be 
filled with pure oxygen which may gradually be allowed to enter 
the system as needed or may be immediately introduced; for we 
have no evidence as yet that atmospheres with an oxygen per- 
centage of 30 per cent or thereabouts alter the metabolism in 
any way. With this method, by simply noting the fall of the 
spirometer and making suitable corrections for changes in the 
barometer and in the temperature, the oxygen consumption is 
readily computed. This method has the disadvantage that the 
oxygjen must be introduced from time to time; the length of the 
period that the apparatus may be left unattended is thus deter- 
mined solely by the size of the spirometer. By attaching a large 
spirometer to the pipe P as, for example, a 50-liter Tissot or 
Bohr spirometer, the oxygen consumption may be determined in 
twenty-four hour periods and the apparatus allowed to operate 
continuously without attention for the entire twenty-four hours. 
With this method we have used a Tissot spirometer which is filled 
at the beginning of the experiment with pure oxygen. 

Since the volume of air in the respiration apparatus is only 30 
Iit<;rs or less, it will be seen that if a spirometer containing 50 
liters of pure oxygen were attached, there would be a rapid 
diffusion of the gas and the percentage of oxygen in the ventilating 
current would soon approximate 50 or 60 per cent. To prevent 
such diffusion, a special form of conmiunication between the 
spirometer and the pipe P has been used, which is shown in Figure 



F. G. Benedict 



307 



2. For this a small S-necked flask has been employed.* Throi^ 
the middle opening {A) connection is made directly with the 
lai^ spirometer; in one of the other openings a glass tube with 
stopcock (B) is inserted to be used for introducing the oxygen 
into the spirometer; in the third opening is a capillary glass nphon 
U-tube (O. 35 cm. long, which connects the suf^y of oxygen 




FiQ. 2. OxygentrapforeonnectinKalargeflpirometarwithtberwpira- 
tion Apparatus. 

with the respiration chamber. The oxygen, being heavier than 
the air in the respiration chamber, is trapped at the bottom of the 
long U-tube and thus is delivered into the chamber only as actu- 
ally required. When the Tissot or Bohr spirometer is used, the 

* Any other form of bottle with a 3-holed rubb«r stopper may be used 
equally well. 



30i8 .. Respiration Apparatus for Small Animals 

S-'^siy valve T^ (Figure 1) is turned so that no air passes into the 
spirometer 0. Since this spirometer is the only flexible part of 
the ventilating apparatus, there will be a slight diminution of 
pres^re' inside the respiration chamber as the oxygen is absorbed. 
This diniitiution in pressure will then be immediately compen- 
sated by an inflow of oxygen from the large Tissot or Bohr spirom- 
eter through the capillary U-tube of the oxygen trap. 

Proof of the efficacy of this method of introducing oxygen 
may be seen in the details of an experiment reported subsequently, 
showing that after a three day experiment with a rabbit the pro- 
portion of oxygen in the air inside the chamber was 20.13 per 
cent. It is clear, therefore, that under these conditions there 
is no appreciable difTusion between the normal air in the chamber 
and the pure oxygen in the spirometer. By noting the differences 
in level of the Tissot or Bohr spirometer, and correcting for 
the changes in temperature of the spirometer bell and the 
changes in the barometer, the oxygen consumption may be readily 
computed. 

Another method of introducing oxygen into the ventilating 
system may be that employed with the large respiration calo- 
rimeters in this laboratory;^ namely, by means of a valve which is 
electrically actuated and connected with the reduction valve on a 
small cylinder of oxygen. Oxygen may thus be automatically 
introduced into the chamber as the bell of the small spirometer 
O falls. By weighing the oxygen cylinder l>efore and after the 
experimental period, the amount of oxygen admitted may be 
measured. 

Finally oxygen may be periodically introduced by allowing 
the gas to flow from a cylinder of oxygen through a carefully 
calibrated gas meter,* such as is commonly used with the universal 
respiration apparatus. 

For our experiments with small animals, we have found the 
method employing the Tissot or Bohr spirometer to be the most 
satisfactory. It is obvious that any type of accurate and readily 
calibrated spirometer may be substituted for either of these 
spirometers; in fact, the compensating devices 6f both the Tissot 

^ See F. O. Benedict and T. M. Carpenter: Carnegie Institution of 
Washington Publications, No. 123, Fig. 31, p. 68, 1910. 

•Benedict: Deutsch. Arch. f. klin. Med.j cvii, p. 181, 1912. 



F. G. Benedict 309 

and Bohr spirometers have no significance in this type of experi- 
ment, as an uncompensated bell may be used with equal accuracy. 

Graphic record of muscular activity. The basal importance of 
knowing the muscular activity of even small animals during 
respiration experiments has hitherto l)een almost completely 
disregarded. Experience in this laboratory has shown that an 
intelligent comparison of experimental periods may not l)e made 
without some graphic indication of the variations in the muscular 
activity, and as far as possible it has been our custom to use for 
this purpose only periods of complete muscular repose. While 
ocular observations of the variations in muscular activity are 
worthless, fortunately the graphic registration of the degree of 
muscular repose is a simple matter. As may be seen in Figure 1, 
the animal is confined in a wire cage suspended by a stout spiral 
spring on one side and by a hook on the other. Under this cage 
is a soft rublxjr bulb (R) which is connected through the wall of 
the respiration chamber with a tambour and kymograph. Each 
change in the center of gravity of the animal's body produces a 
variation in the tension of the air in the bulb; this immediately 
affects the diaphragm of the tambour, whose writing point 
records the movement upon the kymograph. A test of the system 
of recording the muscular activity may be made by gently blow- 
ing into one part of the tube S and noting if the tension of the 
tambour rubber remains constant. By occasionally opening 
the tube at <S, permanent distension of the tambour rubber by a 
major change in the position of the animal may be compensated. 
These graphic records have proved of great value in interpreting 
the results of experiments. 

Method of admihisiering food and drink. In experiments con- 
tinuing several days, it is important to supply water and often 
food to the animal. In many laboratories it is customary to 
open the cage for one hour or less for this purpose and make 
observations of the metabolism only during the remaining twenty- 
three hours; such a method may be employed with this apparatus. 
When continuous experiments are preferred, the method of 
supplying water shown in Figure 1 has been found useful. If by 
means of this device the animal is given a milk diet, such as that 
employed by Laqueur,^ experiments with this respiration apparatus 

' E. Laqueur: Ztachr.f. physiol. Chem., Ixxxiv, p. 109, 1913. 



310 Respiration Apparatus for Small Animals 

may be continued without difficulty for several days, if not weeks. 
The feces pass through a large mesh in the bottom of the cage, 
but are retained by a finer mesh. The urine passes through 
the meshes to the conical section in the bottom of the chamber 
and may be readily withdrawn. 

The details of a t3rpical experiment with a single rabbit, weigh- 
ing 1600 grams, are given in Table I. On the first day of this 
experiment the rabbit was given 300 grams of carrots, and the 
muscular activity incidental to eating the food resulted in a higher 
metabolism; the respiratory quotient of 0.90 is clearly indicative 
of a predominating carbohydrate metabolism. On the subse- 
quent days food was withheld; there was then a continual decrease 

TABLE L 

Metaboliam metuurwierUs with a rabbit {1900 gm.) . 
(No food after first day,) 



CARBON DIOXIDE _> ' 

DAY PRODUCED FEB OXYGEN COK8UKED ■ REBPIRATORT 

24 HM. PEB34HBi. I QUOTIENT 



T 



iBt 26.99 29.87 ■ 0.90 



I 



2d I 18,90 25.85 1 0.73 

3d 16.63 ! 22.32 ! 0.75 

4th , 15.75 21.52 0.73 

5th 14.61 , 20.38 i 0.72 



in the oxygen consumption and carbon dioxide production, this 
being in full conformity with the general picture of the k3miograph 
curves showing the muscular activity of the animal on these days; 
unfortunately the reproduction of these partibular twenty-four 
hour kymograph curves is not feasible. The respiratory quotient 
for the last four days during the fasting period is indicative of a 
predominating fat combustion. 

While the apparatus as previously described is primarily de- 
signed for long experimental periods, a special test was made to 
see what degree of success would be secured with an experiment 
on a much smaller animal and with a shorter experimental period. 
The results are given in Table II. In this test a single guinea pig, 
weighing 400 grams, was used. Instead of the large Tissot 
spirometer the small 2.5 liter spirometer, which is an integral 



312 Respiration Apparatus for Small Animals 

part of thc' unit n.*Kpiratioii apparatus, wa.s employed. Oxygen 
waH admit Uf^l from a previously weighed cylinder of oxygen, 
the .spirometer U*ing filleiJ and the apparatus allowed to run 
unattended until the spirometer IxK'ame nearly empty; thus 
oxygen was admitted periodic&lly. The experimental periods 
wrnt five to seven hours long. 

On the first day of the exfxjriment the guinea pig ate a very 
hirg<* amount of carrots, and the respiratory quotient, 0.94, 
indicates a pn*dominating carlxihydrate combustion. On the 
second day no food was given and the c|Uotient was 0.83, with a 
very marked d(?creasc in the total metaJxilism as measured by 
iUi*. carUni dioxide production and oxygen consumption. On the 
third day the animal was still starving, but another variant was 
introduced in that the temperature of thcj chaml>er was decreased 

TABLE II. 
Mvtahnlimn meuHuremeutH with a single guinea pig il^OOgm.). 

(So food after first experiment.) 

CARBON DIOXIDE 



KXPKItl- 
MKNT 


I'KUIOI) 


PKODrCKD 


OXTOEN CO.N'Sl'MED 


RK8VIKA- 
TOUY 


NO. 




ToUl 


Per hr. 


Toul 


Pit lir. 


yt OTIENT 






laera 


cc. 


liter* 


rr. 




:«> 


\).'M') a.m. 4.37 p.m. 


3.02 


430 


3.21 


4.>7 


0.94 


37 


11.37 a.m. 4.')0 p.m. 


1.45 


270 


1.75 


320 


0.S3 


:w 


0.3,')} a.m. 4.10 p.m. 


1.91 

1 


286 


2.53 


379 


0.75 






1 










_ 



from 2'^A°i\ to 17.3®(.\ Under these conditions the animal was 
v(»ry much more restless and we note in the results a decreasing 
r(»spiratory ciuotient, indicating a predominating fat combustion 
and an incn^ase in the carlion dioxide production and oxygen (con- 
sumption du(» to the activity. The kymograph records for the 
tliree s(»v(»n hour experiments arc reproduced herewith (Figure 3) 
and sliow clearly the influence of muscular activity upon the 
metal)olism and the consequent importance of knowing th(» muscu- 
lar activity wlien int'Crpreting the meta])olism measurements. 

It is not the purpose of this paper to discuss the metabolism of a 
rabbit or a guinea pig in the early stages of fasting or as influenced 
by (H)ld or muscular activity, these experiments being here re- 
ported only as typical cases showing primarily the flexibility of 
the apparatus. 



F. G. Benedict 313 

Metabolism in artificial aimospheres. An apparatus of the 
closed circuit type, such as this, is readily used for studies of the 
metabolism of small animals in artificial atmospheres, and an 
extended research upon the metabolism and pathologic changes 
in small animals while breathing oxygen-rich gaseous mixtures 
is now in progress in this laboratory. In this type of experiment 
the animal remains in an atmosphere containing 90 to 95 per 
cent of oxygen for several days continuously. Under these 
conditions, the carbon dioxide production and oxygen intake have 
been satisfactorily determined. 



312 Respiration Apparatus for Small Animals 

part of the unit respiration apparatus, was oniployed. ()xyfz;en 
was admitted from a previously weighed cylinder of oxygen, 
the spirometer being filled and the apparatus allowed to run 
unattended until the spirometer became nearly empty; thus 
oxygen was admitted periodicJllly. The experimental periods 
were five to seven hours long. 

On the first day of the experiment the guinea pig ate a very 
hirge amount of carrots, and the respiratory quotient, 0.94, 
indicates a predominating carbohydrate combustion. On the 
second day no food w^as given and the quotient was 0.83, with a 
very marked decrease in the total metabolism as measured by 
the carbon dioxide production and oxygen consumption. On the 
third day the animal was still starving, but another variant was 
introduced in that the temperature of the chamber was decreased 

TABLE II. 
Metaholiftin vieasureinents with a single guinea pig i^OOgm.). 

{So food after first experiment.) 



KXPKKI- 
MENT 


PERIOn 






CARBON DIOXIDE 

PRODITCKD 1 

1 


OXTOEN CON8UMEU 


RKBPI K.A.- 
TORY 


NO. 






- 


Total 


' Per hr. 

1 cc. 1 


Total 
liten 


Pit l»r. 
cc. 


QIOTIENT 




Utera 




36 


\)M a.m. 4.37 


p. 


m. 


3.02 


430 


3.21 


457 


0.94 


37 


11.37' a.m. -4.59 


P- 


m. 


1.45 


270 


1.75 


320 


0.83 


38 


0.355 a.m.-4.16 


P 


m. 


1.91 


286 


2.53 


379 


0.75 






_. 





. 








'. .. 



from 23.4°C. to 17.3°C. Under these conditions the animal was 
very much more restless and we note in the results a decreasing 
respiratory quotient, indicating a predominating fat combustion 
and an increase in the carbon dioxide production and oxygen con- 
sumption due to the activity. The kymograph records for the 
thn^e seven hour experiments arc reproduced herewith (Figure 3) 
and show clearly the influence of muscular activity upon the 
metabolism and the consequent importance pf knowing the muscu- 
lar activity when interpreting the metabolism measurements. 

It is not the purpose of this paper to discuss the metabolism of a 
rabbit or a guinea pig in the early stages of fasting or as influenced 
by cold or muscular activity, these experiments being here re- 
ported only as typical cases showing primarily the flexibility of 
the apparatus. 



F. G. Benedict 313 

Metabolism in artificial atmospheres. An apparatus of the 
closed circuit type, such as this, is readily used for studies of the 
metabolism of small animals in artificial atmospheres, and an 
extended research upon the metabolism and pathologic changes 
in small animals while breathing oxygen-rich gaseous mixtures 
is now in progress in this laboratory. In this type of experiment 
the animal remains in an atmosphere containing 90 to 95 per 
cent of oxygen for several days continuously. Under these 
conditions, the carbon dioxide production and oxygen intake have 
been satisfactorily determined. 



SIMPLE QUARTZ MERCURY-VAPOR LAMPS FOR BIO- 
LOGICAL AND PHOTOCHEMICAL INVESTIGATIONS. 

By W. T. BOVIE. 

(From the Laboratory of Plant Physiology, Harvard University, 

Cambridge,) 

(Received for publication, January 18, 1915.) 

TJie quartz mercury-vapor lamps described below were made 
by the writer for use in connection with some experiments on the 
biological and chemical effects of ultraviolet light. At the time 
the lamps were made there were no quartz mercury-vapor lamps 
on the market in this country. The lamps gave, and are still 
giving, after four years' use, excellent results. They are less 
expensive than any of the commercial lamps known to the writer; 
and, furthermore, as will be seen, the design of these "home-made" 
lamps can be altered to suit particular requirements. 

These lamps are not difficult to make. The quartz must be 
worked in an oxyhydrogen flame, and the burner must be ad- 
justed so as to give the highest temperature possible; but the 
technique required for shaping the lamps is simpler than that 
required to make an ordinary T-joint in glass tubing, since the 
quartz joints do not have to be annealed. A good vacuum pump 
is required for exhausting the lamp. The greatest difficulty is 
found in making the seals at the points where the electrodes enter. 
The literature contains a number of more or less complicated 
and impractical methods, but the seals described below are easily 
made and are efficient. 

It may be cheaper at present to have all the quartz parts of the 
lamp made at the factory. As stated above, these lamps were 
made before quartz tubing was being manufactured in this 
country. 

Figure 1 shows a simple form which the lamp may take.* A 
is a quartz tube 1 cm. in diameter. One end is bent downward 

* The form of this lamp is copied from a lamp made by Dr. C. A. Kraus. 

315 



3x6 



Mercury- Vapor Lamps 



to form the positive electrode C. A quartz tube of the siime 
diameter is sealed in, near the other end, to form the negative 
electrode B. The distance between B and C is about 10 cm. The 
end of the tube A forms a condenser. The lamp is sealed, after 
exhausting at H. The tube A should be inclined to the hori- 
zontal position at an angle of 6° to 7°, the end J being higher, so 
that the condensed mercury will flow back into C. Pieces of 5 
mm. quartz tubing are sealed to the lower ends of B and C. These 



W 




Fia. 1. 



in turn are set in mercury cisterns F, F (madp of glass; test-tubes 
may be used). The small tubes are sealed at E with De Khotin- 
sky cement.- The detail of the seal is shown in Figure 2. The 
bulb of wax on the outside is necessary in order to make the seal 
air-tight. The iron wire D conducts the electric current through 
the wax seal. The mercury cisterns F, F radiate heat and also 
make the seal at E more secure by furnishing an additional 
mercury seal. 

2 This cement may be obtained from A. De Khotinsky, 6026 Drexel Ave., 
Chicago, 111. 



W. T. Bovie 



317 



During the process of pumping the lamp should be turned, to 
allow the mercury to flow out of B and C, so as to free any impris- 
oned air bubbles. The lamp should then be placed in an upright 
position and the mercury in the upper parts of B and C boiled 
vigorously, keeping the wax at E cool by having the lamp stand 
in a dish of water. Lastly, the lamp should be connected with a 
110 volt direct current (with suitable resistance in series with it), 
and run for one or two hours. When it appears to run evenly 
the vacuum pump should be stopped, and the lamp sealed off at 
H. The tube A should be drawn down to a small diameter at 
this point before the pumping begins. 



The lamp should be used on a 110 volt direct current with 5 or 
6 sixteen candle power bulbs in multiple as resistance. The 
lamp will carry from 1.25 to 1.5 amperes, depending upon the 
amount of cooling at the electrodes B and C. When convenient 
it is desirable to arrange the experiments so that the tamp can 
stand in a dish of water, with the water surface well up on li and 
C. The lamp is started by tipping, or a Bunsen flame may be 
directed upon the surface of the electrode ('. When the mercurj- 
boils the arc will start. As the lamp is made of quartz there need 
be no fear of breaking by sudden heating oi' cooling. 

The tube A between B and C should be made of transparent 
quartz, but all other parts can be niiKle of the cheaper translu- 
cent, fused silica. A lamp made for physical experiments, where 
only a small Warn of ultraviolet light was needed, was made 



3i8 



Mercury- Vapor Lamps 



entirely of fused silica, save for a small space between B and C, 
about 1 cm. long, where a small piece of transparent quartz was 
inserted. 

Figure 3 shows a convenient arrangement for an experiment in 
photochemistry. The drawing shows a cross-section of the lamp, 
baths, tubes, etc., as used by the writer in studying the tempera- 
ture coeflScient of the coagulation of egg-albumin by ultraviolet 
light. A, A are copper trajrs set in the insulating case B. C is 
the lamp. The quartz test-tubes which hold the albumin are 
supi)orted at D by the shelves E, E. The water in one of the 
trays is heated by the electric heating coil F, while the water in 



o 



fr 




Fia. 3. 



the* other tray is kept at 0°, by placing chopped ice below the 
shelf E. The electric current in the heating coil is controlled by 
a thermoregulator. G, G are stirrers. The partitions H, H serve 
to prevent the stirrers from disturbing the surface of the water, 
and thus causing irregular variations in the reflection of light at 
the surface of the water. In this experiment the mercury cisterns 
(shown at F, F in Figure 1) stood in the ice water bath. 

Figure 4 shows another of the possible forms of the lamp. It 
will be seen that the tube A is vertical. The electrode C, which 
is left open at the bottom, dips into a cistern of mercury M. 
This lamp was designed to run on a 220 volt circuit, and the dis- 
tance betw^een B and C is 30 cm. The lamp, however, will work 
as well on a 110 volt circuit; for, as will be seen presently, the 



W. T. Bovie 



319 



-J 
K 




Fio. 4. 



320 Mercury-Vapor Lamps 

distance between C and B is easily adjusted. The are is started 
in this lamp by elevating the mercury cistern M until the mercury 
in C comes in contact with that in B. The distance between the 
surface of the mercurj'^ in M and the surface in C is equal to the 
height of the barometric column, since the surface at M is exposed 
to the air. When the lamp is running on a 110 volt circuit the 
arc is not stretched down as far as when running on 220 volts. In 
place of the mercury cistern M, one can use a smaller chamber 
closed at the top, with the air space above the mercury connected 
with a water aspirator. The height of the mercury in C can then 
be controlled by varying the air pressure above the mercury in 
the cistern. With this arrangement the distance between M 
and C may be shorter. An air trap L is inserted between M and 
C to prevent the accidental entrance of air bubbles into ^l. 

It will be noticed that the condenser J is not in a line with .1, 
but is a little to one side. This is important, for a certain amount 
(about half) of the condensed mercury must be returned to the 
electrode B, else it will boil dry. The condenser J and the elec- 
trode B are provided with copper radiators, A'. (The radiators 
were taken from discarded Nernst lamps.) By using these 
radiators the lamp could be made to carry a current of four 
amperes. With this current density and with an arc 30 cm. long 
the lamp emits such an abundance of ultraviolet light that the 
air in the room is quickly filled \vith ozone. The photochemical 
power of the lamp may perhaps best be judged from the fact that 
an exposure of one or two minutes at a distance of one meter will 
produce such a ''sunburn" as to cause the skin to peel. Those 
who undertake work with quartz mercury-vapor lamps are warned 
never to look at them ichile they are ninning, except through a screi^n 
of red glass. An exposure of a very few seconds is sufficient to 
cause most painful consequences. The pain does not follow imme- 
diately after the exposure, but there is a latent period of several 
hours, so that one is not aware of the injury at the time of the 
exposure. 



THE METABOLIC RELATIONSHIP OF THE PROTEINS 

TO GLUCOSE. 

By N. W. JANNEY. 
(From the Chemical Laboratory of the Montefiore Home, New York.) 

(Received for publication, January 28, 1915.) 

The importance of glucose as an intermediary metabolic prod- 
uct is constantly becoming more evident. Through "the work 
of Knopf, Glassncr and Pick, Embden, Lusk, Ringer and Dakin,* 
in which depancreatinized or phlorhizinized dogs were used, 
it has been shown that the majority of the amino-acids occurring 
in considerable amounts in the protein molecule are convertible 
into glucose. 

The fact that various proteins yield large amounts of sugar in 
the diabetic phlorhizinized animal has been fully established. 
As the amino-acids, rather than higher complexes, have been 
shown in recent years to be the most important direct products 
of protein digestion, it seems reasonable that glucose derivable 
from catabolizing proteins is formed from the glucogenetic amino- 
acid complexes present in any particular protein. These sugar- 
yielding amino-acids vary markedly in amount in different pro- 
teins.. It follows, therefore, that the proteins in their catabolism 
may be found to yield amounts of sugar proportional to the 
amount of glucogenetic amino-acid radicals present in them. 

Experiments with this end in View have yielded results which 
may thus be summarized. Each protein produces a definite 
amount of ghicose in the phlorhizinized organism. The various 
yields represent 50 to 80 per cent by weight of the protein adminis- 
tered. These yields approximate the ratios which the glucogenetic 
amino-acids of the proteins in each case bear to the total amino- 
acids, as actually determined by hydrolysis. 

» H. D. Dakin : this Journal, xiv, p. 321, 1913. For the literature quoted 
in this article, when not especially noted, reference may be made to the 
comprehensive monograph of Lusk: Phlorhizinglukosurie, Ergebn. d. 
Physiol., xii, p. 315, 1912. 

321 



322 Metabolic Relationship of Proteins to Glucose 

Material. Pure proteins were used in the experiments. Cal- 
culations throughout the paper have been based on anhydrous 
material. Water determinations made by drying in vacuum, 
usually at 110°, have for the sake of brevity been omitted. 

Casein. Prepared according to Hammarsten. 

0.1781 gm. gave 0.0278 gm. nitrogen N = 15.6 per cent. 

According to Hammarsten N= 15.6 per cent. 

Ovalbumin. Prepared from Merck and Go's, ''soluble egg albumin/' 
preparations of which were not entirely soluble in water. They contained 
globulins and ovomucoid. This material is, however, somewhat more 
convenient to use than egg-white. Even with the help of several existing 
descriptions, technical difficulties in the preparation of pure ovalbumin 
may still be met. A jnode of preparation is therefore given here. 

Preparation of ovalbumin. From a' dilute until tered solution of Merck's 
soluble egg albumin or the diluted egg-white, the globulins and ovomucin 
(Eichholz) are removed by adding an equal volume of saturated ammonium 
sulphate solution. The albumin is obtained from the filtrate by dissolving 
in it powdered solid ammonium sulphate until a further addition to a 
filtered sample causes no precipitation. Filter off and dissolve in water. 
Use preferably distilled water throughout. Reprecipitate with ammonium 
sulphate in like manner and repeat this process twice. Collect the albumin 
on a folded filter, wash with saturated ammonium sulphate solution, sus- 
pend the finely divided precipitate in a small volume of water, and dialyze 
for three days to free from the major portion of the sulphate. The albumin 
goes into solution. It may be coagulated by pouring in a fine stream into 
boiling water or by immersing the flask containing it in boiling water. The 
coagulum rubbed fine is washed free from sulphate and ovomucoid by 
decantation or on large folded filters. Ovomucoid is tested for by evapor- 
ating the filtrate to very small bulk and adding four times as much p5 per 
cent or absolute alcohol. A light flocculent precipitate denotes the pres- 
ence of ovomucoid. The pure albumin is dehydrated by letting it stand 
in a finely divided condition for twenty-four hours or longer, under 95 
per cent alcohol, and then washed with absolute alcohol and dry ether 
on a Buchner funnel, with a hardened filter. Damp air should be ex- 
cluded during this process. The product, a snow-white powder, represents 
the total ovalbumin (conalbumin and egg albumin proper). 

0.3652 gm. prepared from Merck's ** soluble egg albumin" gave 

0.05677 gm. nitrogen N = 15.55 per cent. 

0.2005 gm. prepared from egg-white gave 

0.03184 gm. nitrogen X = 15.8 per cent. 

According to Osborne and Harris, 

crystallized egg albumin N = 15.5 per cent. 

conalbumin N=16.1 per cent. 



N. W. Janney 323 

Human serum albumin. Prepared from ascitic fluid from cases of 
hepatic cirrhosis or cardiac insufficiency, according to a method similar 
to the one immediately preceding. Fine white powder. No analyses of 
human serum albumin seem to be reported. The nitrogen values differ 
slightly from that of horse serum albumin. 

1. 0.3484 gm. gave 0.05481 gm. nitrogen N = 15.7 per cent. 

2. 0.3192 gm. gave 0.05033 gm. nitrogen N=15.8 per cent. 

3. 0.2835 gm. gave 0.04458 gm. nitrogen N= 15.7 per cent. 

The analyses were made from different preparations. The material 
for analyses 1 and 2 was three times, that for analysis 3, six times precipi- 
tated with ammonium sulphate. 

Gelatin. A German preparation C*Gold seaF' WH No. 1866). 

0.2994 gm. yielded 0.05186 gm. nitrogen N = 17.3 per cent. 

It was further purified according to the following method: A 20 per cent 
solution in warm water was allowed to solidify, the jelly-like mass finely 
divided by passing through a small meshed sieve, and for a period of two 
weeks washed at a low temperature with distilled water containing ether, 
which was changed daily. The material was then treated successively 
with 50 per cent, 95 per cent, and absolute alcohol and ether. 

0.2867 gm. yielded 0.05076 gm. nitrogen N=17.7 per cent. 

According to Hausmann N= 18.0 per cent. 

Fibrin. A finely pulverized preparation obtained from blood. 

0.3965 gm. gave 0.06577 gm. nitrogen N = 16.6 per cent. 

According to Samuely N = 16.9 per cent. 

The vegetable proteins were prepared according to Osborne.* 
Edestin. Prepared from hemp seeds. 

0.2718 gm. gave 0.05115 gm. nitrogen N = 18.8 per cent. • 

According to Osborne N « 18.6 per cent. 

Gliadin. Prepared from wheat flour. 

0.2520 gm. gave 0.04429 gm. nitrogen N=17.6 per cent. 

According to Osborne N = 17.7 per cent. 

Zein. Prepared from corn gluten meal. 

0.3133 gm. gave 0.0506 gm. nitrogen N = 16.2 per cent. 

According to Osborne and Harris N = 16.1 per cent. 

Mode of feeding. Casein, ovalbumin, scrum albumin, fibrin, and ede* 
stin were thoroughly moistened with water, and when not voluntarily 
eaten were fed through the stomach tube. Gelatin was prepared for ad- 
ministration by dissolving in about five times its weight of warm water. 
The jelly formed on cooling was finely divided and then fed, or the gelatin 
solution at 40° was introduced into the stomach by the stomach tube. 

Gliadin when moistened with water or saliva becomes a dense glue- 



* T. B. Osborne : Darstellung der Proteine der Pflanzonwelt, in E. Abdor- 
halden: Handbuch der biochemischen Arbeitsmeihoden, ii, p. 270, 1910. 



PKRIOUB 


; NITROQEK 


GLUCOSE 


o:n 


hra. 


1 om. 


- ■ i- 

gm. 




J '2 


3.77 


13.11 , 


3.48 


I 12 


3.35 


11.76 


3.61 


24 


i 7.44 


26.09 ! 


3.50 



324 Metabolic Relationship of Proteins to Glucose 

like mass, exceedingly difficult to handle or to swallow. A special feeding 
method was therefore devised as follows: The gliadin was dissolved in 
a^ small an amount as possible of j^ sodium hydroxide, and the solution 
was introduced into the stomach by the stomach tube, followed immedi- 
ately thereafter by an equivalent amount of /jy HCl. The gliadin was 
thus deposited in a very soft digestible condition in the stomach. 

Alkali-acid control experiment. 



Weight of dog, 9.2 kgm. 384 cc. 
.^^ NaOH followed by 384 cc. ^^' 
HCl (80 cc. fluid \yer kgm.) by' 2 
stomach tube given on 2d day.i 

__ _L _• j_ _ ' _ ' _ . 

The technique employed in this experiment was precisely the same as 
in the feeding experiments. Neither digestive disturbances nor influence 
on the glucose excretion was observed. The similarity of the gliadin 
partition curve with that of the proteins fed without the use of acid and 
alkali demonstrates also the reliability of this method of feeding. Very 
large amounts of less dilute alkalies and acids fed in this manner have, 
however, been shown by other experiments to give rise to diarrhea, diuresis, 
and even to metabolic disturbances giving rise to an increased glucose 
excretion. 

Zcin, on account of its hard consistence, is difficult of digestion and 
absorption, as is well known. Thus, in a recent paper of Osborne and 
Mendel,' reporting a comparative series of feeding experiments with pro- 
teins, the utilization of zein (average of seven experiments) was 65 per 
cent for the rat. This is considerably less than in the case of other pro- 
teins. For this reason a special method of preparing the zein for feeding 
was devised as follows: 

Finely divided zein is shaken with sufficient ^6 sodium hydroxide at 
room temperature to obtain a 5 per cent solution. This solution, which 
must not be allowed to stand longer than twenty-four hours, is precipitated 
by adding, while stirring, a like volume of -^ hydrochloric acid in a fine 
stream. The soft mass of zein is freed from excess of the neutral fluid 
by draining with suction on a Buchner funnel and then given to the animal 
through the stomach tube immediately after adding sufficient water to 
make up the standard amount required for the experiment. The filtrate 
was evaporated to dryness at 40**. The residue containing zein was then 
treated with absolute alcohol to take up tfce protein, the solution filtered 
and evaporated to dryness again at a low temperature, and finally treated 
with water to free it from traces of sodium chloride. The pure zein thus 



» T. B. Osborne and L. B. Mendel: this Journal^ xviii, p. 177, 1914. 



N. W. Janney 325 

obtained on drying usually amounted to about 0.15 gm. This amount 
was deducted from that weighed off for feeding. By the use of this method, 
:us the protocols show, zein can be fed with certainty of obtaining as com- 
plete absorption as ia the case of any other protein of this scries. 

A standard amount of protein containing usually 0.3 or 0.5 gm. nitro- 
gen per kgm. was fed, at times with the help of a little dilute extract of 
beef. Extract of beef has been shown by Lusk not to influence the glucose 
elimination. 

In view of the necessary emplo>iiient of the stomach tube for feeding 
in many cases, the question of psychic glycosuria with consequent influ- 
ence on the sugar excretion may be raised. By comparing the results, 
however, of experiments where the protein was eaten voluntarily with 
th<»f*e where the tube had to be employed, we have never in a consider- 
able experience seen results where a possible extra sugar excretion due to 
struggling could be detected. Hirsch and Reinbach^ report that even con- 
tinued and violent excit<»ment in dogs is followed by but a relatively 
small increase of blood sugar. Even in this case the urine containe<l 
IK) ^lucose. 

Methods. Two days previous to urine collection fasting began, and 1 gm. 
of phlorhizin was administered according to Coolen daily, in some cases 
every twelve hours. Periods of urine collection were separated by cathe- 
terization and washing out the bladder with aseptic precautions. The 
last washing terminated precisely with the period's end. Nitrogen was 
detennined according to Kjeldahl, glucose according to the Pavy-Fehling 
method, and a polariscopic control was made in all cases. In certain of 
the casein experiments glucose was titrated according to Benedict.* 

Much more care is nniuired in the choice of dogs for the ex- 
l>eriments here reported tlum is necessary for those aiming mere- 
ly to demonstrate qualitatively a rise in the sugar elimination 
after ingestion of a given substance. Only hardy mature ani- 
mals, preferably about two years old, whose general tone remains 
excellent during the entire experimental period, can l)e relied upon 
for accurate results. A weakly or young dog may excrete but 
75 or even 50 per cent of the maximal yield of glucose obtainable 
from a given protein. In properly selected animals, however, 
if due care was observed, a surprising regularity in the amount of 
the glucose yields of different experiments was not(*d. 

The amounts of protein fed were so chosen that the glucose yield even 
in larger dogs fell below 25 gm., with but one exception. Phlorhizinized 

* E. Ilirsch and H. Reinbach: Zlschr.f. physiol. Chem., xci, p. 29J, 1914. 

* For these experiments and for a preparation of gliadin and of edestin, 
I am indebted to Dr. Isidor Ctreenwald, now of the Harriman Research 
Laboratory, New York. 



326 Metabolic Relationship of Proteins to Glucose 

dogs excrete quantitatively glucose ingested in amounts at least up to 
50 gm. or more.* For the calculation of the extra glucose the method of 
Lusk was adhered to. Though wide differences in the amounts of body 
protein spared by the ingested proteins were observed, they influenced 
but slightly the calculation of the extra glucose. Onl}' one experiment 
with a very irregular nitrogen excretion did not lend itself toLusk's method 
of calculation. 

PART I, THE RATE OF BiETABOUSM OF PROTEINS. 

In the present study it is of fundamental importance that the 
proteins be administered in such a way as to insm*e maximal 
absorption and prompt elimination of their metabolic products. 
If these conditions are not secured the results obtained would have 
but little value. Mendel and coworkers^ have carried out re- 
cent studies in this direction, the general conclusion reached 
being as follows: it is probable that the digestive products of 
all proteins are absorbed and eliminated at like speed provided 
that certain factors are allowed for. The latter are: 

(1) Mechanical. A very finely divided, thoroughly softened, 
and hydrated prot'Cin in a pure state is digested and absorbed witli 
great rapidity. Admixture of indigestible substances, such as 
cellulose, delays these processes in direct proportion to the rel- 
ative quantity present. 

(2) The amount of carbohydrate and fat as protein sparers, in 
the diet. 

(3) Wat^r intake with the meal.* The differences of previously 
reported experiments dealing with absorption and utilization of 
proteins can be accounted for on these grounds. With proper 
allowance for the factors mentioned, rapid and complete or nearly 
complete absorption of most proteins can be obtained. 

•A. I. Ringer: Proc, Soc, Exper. BioL and Med,, ix, p. 52, 1911-12; 
this Journal^ xii, p. 431, 1912. See also Lusk: loc. cit. 

^ For the literature referred to in this section, reference may be made, 
when not especially noted, to K. P. Cathcart: The Physiology of Protein 
Metabolifimf New York, 1912. L. li. Mendel and M. S. Fine: this Jour- 
nal, X, pp. 3a3, 339. 345, 1911-12; xi, pp. 1 and 5, 1912. L. B. Mendel 
andR. C. Lewis: idem., xvi, pp. 19, 37,55, 1913-14. 

• This is important as it has recently been demonstrated that water 
is an active gastric stimulant (O. Bergeim, M.E. Rehfuss, and P. B. Hawk: 
this Jouninl, xix, p. 345, 1914). 



N. W. Janney 327 

In the following experiments these studies have been extended 
to the phlorhizinized dog. Proteins were fed and hourly deter- 
minations of the glucose and nitrogen made. With a view of 
obtaining complete absorption, all the determining factors as 
above mentioned are given, it is thought, due consideration. In 
addition, it seemed wise to employ fasting animals as well as small- 
er amounts of proteins for feeding than have been generally 
employed hitherto. This latter was done because the writer has 
observed that the amounts of protein food large enough to im- 
pede complete absorption are much smaller than is generally 
supposed. No feces were passed in any of the experiments. 

The curves show the percentage of the extra nitrogen and 
sugar excreted per hour. In each case the mean hourly nitrogen 
and glucose excretion was obtained from the fore- and aft^r- 
periods immediately adjoining the nine hourly periods. The 
excretion in excess of this mean for these periods was then noted, 
and the sum taken as 100. The glucose and nitrogen curves for 
one protein show no extraordinary deviation from those obtained 
for the others. In each case the glucose excretion reaches its 
maximum at the second to the third hour, while the highest 
hourly excretion of the nitrogen is usually attained a little later, 
from the third to the fifth hour. Ehmination of the glucose and 
nitrogen is complete by the ninth hour in all cases. An identi- 
cally planned partition experiment was also carried out in the 
case of pure zein. As the experiment was imperfect it is not re- 
ported in detail. The glucose was completely eliminated by the 
ninth hour, the resulting curve being practically identical with 
those described in this article. The glucose and nitrogen curves* 
in the case of serum albumin are more flattened than any of the 
others. The probable explanation for this lies in the fact that in 
this experiment but little water was given with the protein (page 
326). The two casein experiments are of interest as their nitrogen 
curves present three waves. A similar wave-like nitrogen excre- 
tion has been reported by others.''^ An entirely satisfactory' 
explanation of this seems lacking. 

* The author is indebted to Mr. Frank C. ( -sonka for this experiment. 

'»B. Tschlcnofif: Ceutralbl. f. Physiol., x, p. 177, 1896-97. 0. Veragiith: 
Jour, Phyaiol.f xxi, p. 112, 1897. E. Haas: Biochein. Ztschr., xii, p. 203, 
1908. 



328 Metabolic Relationship of Proteins to Glucose 




Chart I. Casein I. In the cuncs height roprcw-ntB glucoee and nitro* 
gen hourly percentage- cxuretion; lenfith, hourly periods. The dotted curve 
represents glucow (•xrretion, mid the solid curve the nitrogen. 




N. W. Janney 





^ 


^, 


i* 






;^^ 


^ 


i 


it 


+H---±W; 


15 


— 


iz. 


xii 




# 






V 




tl-L 


1 


10 


-hr 


ii-" 


■± 


44^ 


^ 




N 


'II' 


~~ 


± ^ 


5 




:rr,-x 


-S--ZZ 






— ;-.-!- 




^ 


£ 






















/ -"t 


— l-L 


— r— 


-p}--^ 


— [-j-^ 


H-T-^-- 


r^'^ 




i\ 











































12345678 9 

Chart 3. Serum olbur 



3( 
25 


h I 


-h-p^ 


^ 




II- :^-^ 


-M-f-- 


rH" — ^ 


IS 
10 










4 


y 




4 


r-r-j-j ■ 


TT : 


4r 






4t^ 


Sp±iH 




-JT^ 


r^^tr^ 


-44+^ 


tfflrrr 


^1:# 





1 


^ 


jmi 


4 


#^ 


7 


-\— f. 

8 9 



Chart 4. Gliadin. 



33° Metabolic Relationship of Proteins to Glucose 



55. |||[|||||| III- iiiiiiiiii'ir niiMiiiiH 







































































3 + 5"^ 

Chabto. Edestin. 




I 2. 3 4 5 6 7 8 

CiiAHT G. Combined curve of the five pro.tcin experiments. 



N. W. Janney 331 

The nitrogen curves obtained in these experiments do not 
differ greatly from those reported by others for normal dogs, 
though but few of the many protein feeding experiments pre- 
viously published have been made under very similar experi- 
mental conditions. The work of Feder, Graflfenberger, Stauber, 
Wolf, Mendel and Lewis, and others may be referred to. It is 
evident that the phlorhizinized kidneys are still capable of per- 
forming their normal function of excreting the end-products of 
nitrogen metabolism. 

It seems remarkable that the glucose curve in most cases 
reaches its maximum before the nitrogen curve. An explanation 
for this may be in the fact that the nitrogenous portion of the 
protein is catabolized to its normal end products, while the non- 
nitrogenous has its metabolism arrested at the stage of glucose, 
an intermediary product. It is possible, however, that the greater 
permeabiUty of the phlorhizinized kidney for glucose may be 
the true explanation of this phenomenon. 

The digestion of proteins is seen to be, if we take the glucose 
cur\'e as an index, still more rapid than the nitrogen curve would 
indicate. With figures which represent the mean of all five 
partition experiments here reported, 35 per cent of the extra 
glucose was found to be eliminated by the end of the second 
hour against 18 per cent of the nitrogen. 

The glucose nitrogen ratio becomes very much increased in 
the first few hours, due not only to the greatly increased excre- 
tion of glucose, but also to the lack of coincidence of the glucose 
and nitrogen curves. The gliadin experiment well illustrates 
the fact that the size of the G :N ratio of the experimental period 
depends not only on the nitrogen and glucose formed from the 
ingested substance, but also on the time included in this period 
before and after the excretion of the extra nitrogen and glucose 
has begun and ended. This has been frequently overlooked in 
earlier work done on phlorhizin diabetes. It is illustrated by 
the following table from the gliadin partition. Time, as noted, 
l)egin8 with the ingestion of the protein. Mean G:N of the fore- 
and after-periods is 3.20. The calculated extra sugar remains, 
however, about the same for each period. 



332 Metabolic Relationship of Proteins to Glucose 



PKRIODB 


u:n ratio 


KXTRA OLUCO.iE 


hrs. 


1 




9 


4.09 


86.7 


12 


3.88 


84.8 


24 


3.65 


84.7 



The prcsont partitions seem to be the only ones on re^cord 
where isolated proteins have been administered to fasting plilor- 
hizinized dogs. Similarly planned experiments have, however, 
been carried out by Reilly, Nolan, and Lusk" in three experi- 
ments with feeding of meat. The results obtained are essentially 
the same as here reported, though the nitrogen and glucose of 
the meat were excreted somewhat more slowly. This is probably 
due to the relatively larger amounts fed. 

A uniform plan has been adhered to in carrying out the above 
experiments. The curves of the nitrogen and glucose are similar 
in all cases, and represent the time taken for excretion of the 
products of both the nitrogenous and non-nitrogenous portions 
of the ingested protein. ThiSj it is believed, tnakes it clear thnt (he 
rate of metabolism in the animal organism. ?«, xuider optimal con- 
ditions, alike for both vegetable and aninml proteins. 

Proteolytic enzyme action is mainly a matter of simple hydroly- 
sis. It w'ould seem improbable, a priori, that a difference in the 
l)ehavior of pure proteins introduced into the body in a highly 
jussimilable condition, whether of animal or vegetable origin, 
should evince itself. Vegetable proteins ingested in an impure 
state with cellulose, fiber, and other impurities are, indeed, as 
experience has shown, often less digestible than are animal pro- 
teins of softer texture. That this difference of digestibility does 
not depend upon chemical structure of the protein, the work 
here reported seems clearly to indicate. It must be emphasized, 
however, that the experiments here presented were made in the 
presence of the urgent need of the body for protein food and with 
excess of enzymes. It must be admitted, then, as possible that 
differences in the digestion rate of various proteins may occur 
under less favorable conditions. 

" F. H. Reilly, F. W. Nolan, and G. Lusk: Am. Jour. Physiol., i. p. H95, 
1.S9S. 



N. W. Janney 333 

PART II. METABOLISM OF PROTEINS IN THE PULORHIZINIZED 
ORGANISM AND RELATED QUESTIONS. 

The amounts of extra glucose yielded by a series of eight pro- 
teins has been determined in the following experiments. They 
have been carried out under the same plan as that of those al- 
ready described, except that twelve or twenty-four hour periods 
were employed. In view of the data already presented it seems 
certain that all the extra glucose and nitrogen were eliminated 
well within these periods of time. 

The reliability of the results of these and other experiments 
in phlorhizin diabetes depends upon accepting the statement 
that the extra sugar originates from the substance ingested. 
Direct proof of this is lacking. It seems, therefore, not out of 
place to discuss this subject before describing the results of the 
experiments. 

Origin of the extra glucose in phlorhizin diabetes. 

For the rise in glucose elimination following ingestion of various 
substances by phlorhizinized animals, there can be but the follow- 
ing possible sources. The extra glucose has its origin either 
in the carbohydrates, fats, or proteins of the body, or in the 
ingested substance. 

All previous experimental evidence demonstrates the rapid 
removal of carbohydrates at the beginning of the fasting phlor- 
hizin period. The relation of the body glycogen to this question 
has been fully discussed by Lusk. It may suffice to state that 
although a small amount of glycogen and possibly of other carbo- 
hydrates persistently remains in the tissues of the fully phlor- 
hizinized fasting animal,^- no evidence has been advanced demon- 
strating that the ingestion of small portions of glucose-yielding sub- 
stances can either lead to an increase in, or an elimination of, these 
carbohydrate rests. The breakdown of fat is without influence on 
the sugar excretion (Lusk and others). During the entire phlorhizin 
starvation period, breakdown of body protein leads to the elimina- 

** 0.06 per cent liver (Pfliiger and Junkersdorpflf). According to un- 
published analyses made by Mr. Frank C. Csonka in this laboratory, the 
muscles of fasting phlorhizinized dogs contained 0.03 per cent glycogen 
and 0.56 per cent total substances reducing Fehling's solution. 

TBS JOUSNAI. or BIOLOGICAL CBSSflSTRT, VOL. XX. KG. 3 



334 Metabolic Relationship of Proteins to Glucose 

tion of the resulting glucose in definite ratio with a corresponding 
amount of nitrogen. It seems reasonable, therefore, to suppose 
that the corresponding amount of nitrogen would be eliminated 
should the extra sugar appearing after the ingestion of a given 
substance be derived from body protein. A long series of experi- 
ments is on record in which the ingested substance, which itself 
contained no nitrogen, led to extra sugar formation without any 
appreciable rise in nitrogen elimination. These experiments 
may be induced as indirect evidence against the origin of the extra 
glucose from body protein. 

In support of the view that the extra glucose is formed from the 
. ingested substance, the following may be mentioned. (1) (jIu- 
cose fed to phlorhizinized dogs in the same amounts as the extra 
sugar excreted after feeding of glucogenetic substances, reappears 
quantitatively in the urine. It is thus made probable that glu- 
cose arising in the organism in like amount would also be elimi- 
nated quantitatively. (2) In no case, in the author's knowl- 
edge, roughness of calculation of the extra sugar being considered, 
has the extra sugar found been greater in amount than that de- 
rivable from the carbon of the compound ingested. (3) Ad- 
ministration of substances, incapable in view of their chemical 
structure of being converted into glucose, has l>een found not to 
be followed by extra glucose formation. (4) In the case of 
ingested proteins and other glucogenetic substances containing 
nitrogen, the extra glucose elimination rises and falls nearly 
parallel with an extra nitrogen elimination, as shown in partition 
experiments; from which it may be inferred that the origin of 
this extra glucose and nitrogen is in the ingested substance. (5) 
Evidence showing that the extra sugar derives its origin from 
other sources than the substance ingested has not been advanced: 
and, indeed, data tending to disprove that the origin of the extra 
sugar is in the ingested substance seems also to be lacking. 

Glucose yields of ingested proteins. 

f I • a 

I'UOTEIN ! -* I « ._ 21 P 



fa ' *5 ^ D 

< ■. ?r ^ , K 

w O OS 



Glucose yield in per , | 
cent 48 54 55 , 65 53 05 80 53 



'A 


1 


^ 


t. 


S 


' 2 




B 


< 


1 S 


a 


^ 


H 


n 


o 


•J ■ 


C 


i E 


u 

05 


o 


55 


1 
p 

53 


80 ' 



n 

M 



N. W. Janney 335 

The percentages of protein converted by the organism into 
glucose represent the average of from four to eight experiments 
in each case. The results of the individual experiments did not 
show very wide variations. The amounts of glucose derivable 
from the proteins examined are seen to represent from one-half to 
four-fifths of the weight of the invested protein. From the stand- 
jmnt of amount at lea^t, dextrose mv^t thv^ be considered one of the 
chief intermediary products of protein metabolismy and of certain 
proteins the most important. 

Results of previous experiments. 

Of similar experiments already on record, but few may safely 
Ik* used as a basis for calculation of extra sugar yields. The 
second of two gelatin experiments by Lusk" has been recalculated 
I)recisely according to the method employed in the experiments 
reported in this article. The yield obtained is 64 per cent. 
Halsey obtained yields of 49 per cent for casein and 50 per cent 
( recalculated) for fibrin. Lower glucose values in other of Hal- 
«!iy's casein experiments are probably due to the relatively largo 
amounts of this material fed. 

delation of the chemical constitution of proteins to their glucose 

yields in the phlorhizinized organism. 

In view of the studies of Embden, Lusk, Dakin, and others' 
the amino-acids of proteins may be classified as to their behavior 
in the glycosuric organism according to the following scheme." 



Amino-acids. 


GLUCOGENETIC 


NON-GLUCOOENETIC 


Glycine 


Valine 


Alanine 


Leucine 


Serine 


Isolcucine 


Cystine 


Lysine 


Aspartic acid 


Phenylalanine 


Glutaminic acid 


Tyrosine 


Ornithine 


Tryptophane 


Proline 





Of the non-glucogenetic amino-acids, leucine, tyrosine, and 
phenylalanine yield an increase of aceto-acetic acid (Baer and 

**Reilly, Nolan, and Lusk: loc. cit. 
** Dakin: loc. cit. See table. 



336 Metabolic Relationship of Proteins to Glucose 

Blum, Ringer and Lusk, Dakin). The remainder of this group 
have not been dearly demonstrated to yield either glucose or 
aceto-acetic acid. 

Ingested protein has been shown to break down rapidly into 
its component amino-acids; so it may be said that the metabolism 
of protein and amino-acids is, after this stage, practically the 
same. One might, therefore, expect to be able to calculate fairly 
definitely the amount of glucose a given protein should yield. 
Various reasons render this impossible. The analyses of hydro- 
lyzed proteins are vciy incomplete, as only about 48 per cent, ac- 
cording to Osborne, of the amino-acids can be actually determined. 
Even if this gieat source of error could be eliminated, the sum of 
the amino-acids obtained from a given protein would not represent 
the exact amount of the protein analyzed; for they contain, 
in addition, water acquired by hydrolysis. Again, all proteins 
are not composed entirely of linked amino-acids. These acids 
have, moreover, Ix^en shown to yield amounts of glucose in the 
diabetic organism, which vary widely, from 93 per cent (alanine) 
to 62 per cent (aspartic acid) by weight. 

An attempt, however, has been made to establish a basis of 
comparison. The results obtained with the method used are 
certainly of comparative value. The calculation is made as fol- 
lows: As determined bv the usual methods the amino-acids 
form, in general, a fairly representative fraction of the total 
amount of these acids present in the protein. There are found, 
of course, larger percentages of some than of others. As this 
is true for the glucogenetic as well as the non-glucogenetic 
acids, a certain, if very rough, balancing of this source of error 
evidently exists. The percentage relationship of the glucogenetic 
amino-acids with the total amino-acids as determined by analyses 
approximates that existing between the glucogenetic portion of 
the protein molecule and the entire molecule. The ratios so 
obtained have been gi*aphically represented in the following 
figure, and for the sake of comparison the actual yields of glucose 
in per cent have been drawn to the same scale. 

Though discrepancies occur, as would be expected from the 
rough mode of calculation, the resxilLs as tabulated demonstrate in 
rnost case,s a. definite correspondence between the calculated glucose 
ratios and the value.^ found by direct experiment. When the total 



N. W. Janney 337 

amounts of amino-acids as obtained by hydrolysis represent two- 
thirds or more of the protein, as is the case for casein, edeslin, gliadin, 
and lein, this correspondence is very close, less than 5 per cent ran'o- 
tioii being observed. 




Fio. 1. Glucose derived from protein. 



For the cases of widest variation from the calculated valuea, it 
is probable that the great incompleteness of the analyses is re- 
sponsible. Thus, gelatin, which is peculiar in containing no tyro- 
sine nor tryptophane, much proline, and glycine in greater abund- 
ance than almost any other protein, may be expected to yield 
considerable glucose. The high calculated ratio {see Figure 1) 
ia, however, evidently due to the relative preponderance of gly- 



338 Metabolic Relationship of Proteins to Glucose 

cine and proline in the fraction (45 per cent) of the araino-acids 
of gelatin accounted for by hydrolysis. 

Gliadin forms more sugar in the organism than any protein 
examined. Fortunately, in this case, the amino-acids as de- 
termined by hydrolysis show a total of 84 per cent, which makes 
more definite calculations possible. 

Analysis of gliadin^ 100 grams. 

• 

I i 

ULUCOOKNETIC ASIINO- . ANALYSIS OF PROBABLE ACTUAL I CALCULATED 

ACICS PBS8ENT | OSBORNB | AlfOUNTB PBX8KNT I OLUCOSK TIBU> 

Alanine ' 2.0 ' 4.3 4.0 

Serine I 0.2 3.0 j 2.6 

Cystine... 0.5 1.0 0.7 

Aspartic acid ' 0.6 1.4 0.9 

Glutaminic acid | 43.7 63.3 42.1 

Proline 13.2 21.7 



Arsinine 3.2 3.5 

^ I I 



15.8 
2.0 



Total I 63.4 i 98.2 68.1 

The attempt has been made to calculate the probable actual 
amounts of the glucose-yielding amino-acids present in gUadin.'* 
They total to nearly 100 per cent (column 2). This is, however, 
by no means an impossible sum; for the addition of water to the 
amino-acid radicals of the protein on its hydrolysis greatly in- 
creases the combined weight of the hydrolytic products. From 
these values the glucose yields (column 3) have been computed, 
using in each case the average glucose yield of all the experiments 
reported in which the amino-acids in question have been fed to 
fasting phlorhizinized dogs. The amounts of glucose so cal- 
culated total 68 per cent. By the comparative method descril>ed 
above the calculated glucose is 76 per cent. The actual averajj^e 
sugar yield obtained from gliadin by direct experimentation is 
80 per cent. Taking into account the rough mode of calcu- 
lation necessarily employed, the results in all cases show no 
greater deviation from each other than might be expected. The 
yield of glucose from protein undergoing normal digestion might 

*» For the data employed see T. B. Osborne and D. B. Jones: Am. Jour. 
Physiol., xxvi, p. 305, 1910. D. D. Van Slyke: this Journal, ix, pp. 185 and 
205, 1911; X, p. 15, 1911-12. Cathcart: loc. cit. 



N. W, Janney 



339 



fairly be expected to be relatively higher than that from isolated 
amino-acids, some of which are distinctly toxic when introduced 
into the organism in any considerable amount. For further 
calculations the following data have been employed : 

According to Osborne, 

100 gm. gliadin C=o2.7 gm. H = 6.9 gm. = 21.7 gm. 

N=17.7 gm. S=1.0 per cent. 
80 gm. glucose C=32.0 gm. H=5.4 gm. 0=42.6 gm. 

It has been calculated that 61 per cent of the carbon of gliadin 
leaves the body in the form of glucose. Judging from the oxygen 
contained in 100 grams of gliadin, 22 grams, and that contained in 
80 grams of glucose, 43 grams, it would seem impossible that 80 
grams of glucose could be formed from 100 grams of gliadin. This 
additional oxygen may, however, be accounted for by the process 
of hydrolysis, and oxygen may further be added in the process of 
synthesis of glucose from the products of protein digestion. 100 
grams of the elemental components of gliadin may be approxi- 
mately accounted for as follows: 59 grams of carbon, hydrogen, and 
oxygen combined in glucose, 18 grams as urinary nitrogen. The 
remaining 23 grams as yet unaccounted for consist chiefly of leu- 
cine, tyrosine, and phenyalanine, which yield no glucose. But aceto- 
acetic acid in the diabetic organism totals 12 percent of the en- 
tire amount of amino-acids obtained from gliadin. It seems 
probable, then, that some of the residual carbon may go to form 
aceto-acetic acid and related bodies. 



The ratio glucose: nitrogen of proteins. 



PBOTBIN 



G: N ratio 







6 n 




IBBIN 


s 


LIADIN 


s 


o 


m 


O 


K 


H 


O 


3.08 


3.49 


ZA% 


3.59 


3.16 


3.48 


4.53 






3.29 



The above table has been calculated from the percentage of 
nitrogen in the proteins and the glucose yielded by them in the 
phlorhizinized organism. With the single exception of gliadin, 
it is seen that the G :N ratios thus calculated vary to no greater 
extent than do the urinary G:N ratios of any extended series 
of fasting phlorhizinized dog experiments. The average ratio 



340 Metabolic Relationship of Proteins to Glucose 

for fibrin, gelatin, and serum albumin, all proteins of wide dis- 
tribution in the adult animal, is 3.41, — considerably lower than 
the G:N ratio for body protein, 3.67, established by Lusk. No 
certain conclusion can, however, be drawn from results repre- 
senting but three proteins of the great number present in the 
animal body. 

Protein minimum and growth cormderations. 

The interesting work of Michaud^* and Zisterer^^ has shown that 
the least quantity of protein (the protein minimum) which must be 
digested in order to prevent loss of body protein varies according 
to the nature of the protein fed. The protein minimum is least for 
those proteins nearest in composition to '* Korpereigenes Eiweiss.'^ 
In order to make a comparison possible between the size of 
Zisterer's protein minima for the dog and the glucose yielded by 
ingested proteins in the same animal, when phlorhizinized, these 
values for casein have been taken as units, and the following 
graphic representation has been made. 

The glucose yields are seen to correspond roughly to the size 
of the protein minima. The latter in the case of casein is less 
than that of the other proteins. The glucose produced in the 
organism by this protein approximates that of dog, fish, and 
ox muscle.** From these data the inference may be drawn that 
edestin and, in still greater degree, gliadin contain a balance 
of the various amino-acids less suitable for utilization in the dog 
than that found in casein, and that their glucose yields in the 
diabetic animal may be taken as an index of this unfavorable 
constitution. 

But the problem of the nutritive value of food proteins is more 
complex than the above would indicate. The recent important 
work of Osborne and MendeP' has demonstrated in the most 
convincing manner that both maintenance of body weight and 
growth in animal depend directly on specific amino-acids, tryp- 

" L. Michaud: Ztschr,f. phyaiol. Chem.y lix, p. 405, 1909. 

>M. Zisterer: Ztschr. f, Biol., liii, p. 157, 1909. 

'• Unreported experiments. 

•'T. B. Osborne, L. B. Mendel, and E. L. Ferry: Ztschr. f. physiol, 
Cfiem., Ixxx, p. 307, 1912; this Journal^ xvii, p. 325, 1914 (where reference 
to other papers is made). 



N. W. janney 



341 



tophanc being neceSBary for maintenance of body weight and 
lysine for growth. Neither of these amino-acids yield glucose. 
The glucose yields of the proteins in the diabetic organism could 
not, then, be expected to bear any close relationship to growth 
problems. Casein, ovalbumin, and zein yield amounts of extra 
glucose which do not vary greatly from one another. Casein 
and ovalbumin are rich in lysine and promote maximal growth, 
while zein fails to do so on account of its lack of lysine and tryp- 
tophane. The conclusion may be drawn from these and other 




examples that the amounts of glucose yielded by the metabolism 
of proteins stand in no obvious relationship to their ability to 
promote growth. 



The sparing of body protein effected by the ingested proteins. 

A detailed study of this subject is beyond the scope of the 
present article. The method of calculation of the body protein 
spared by the ingested proteins was the usual one. From the 
average nitrogen excretion for the immediately preceding and 
succeeding periods, the amount of nitrogen eliminated in the 
experimental period in excess of that contained in the ingested 
substance was deducted. The sum so obtained, expressed in 
per cent of the average nitrogen excretion, is accepted as the per 
cent of body protein spared by the fed protein. Wide variations 
in the amount of body protein spared in different experiments 



342 Metabolic Relationship of Proteins to Glucose 

were observed. The following table records the average results 
for each series of experiments. 



CASEIN' 



41 



OV ALBU- 
MIN 



39 



BERUIC 
ALBDICIN 



33 



GELATIN FIBRIN ' BDBSTIN ai.IADIN 



I- 



22 



27 



16.5 



26 



IF.IN 



33 



• Casein and ovalbumin spare somewhat more protein than the 
other proteins, possibly because their composition may l)e more 
adapted to the general requirements of the body fqr amino-acids 
(Figure II). No marked difference is not^d in the ability of 
the vegetable and of the animal proteins to spare body protein. 
In an earlier part of this paper it was shown that no difference 
was demonstrable in the digestibility of proteins of animal or 
vegetable origin. These facts stand in accord. 

Relationship of protein to carbohydrate metabolism. 

The importance of glucose as an intermediary product of 
protein metabolism is clearly shown by the data presented in 
this article. Fifty to 80 per cent of the ingested proteins and 
about 59 per cent of body proteins (Lusk) are converted into 
glucose in the phlorhizinized organism. A majority of the amino- 
acids occurring in proteins goes over into glucose. This forma- 
tion of large amounts of glucose in protein metabolism may be 
considered a normal process.^^ 

The important protein-sparing function of carbohydrate is 
known. It is also a well demonstrated fact that utihzation of 
protein in some manner practically depends upon the presence 
of cai'bohydrate. The experimental work dealing with this sub- 
ject has been collected and ably presented by Cathcart.^* Dakin 
has shown that oxyaldehydes are probably very important me- 
tabolic substanc<?s related closely to glucose and to amino-acids 
as well. It could well be that glucose and amino-acids are 
broken down to oxyaldehydes, which add ammonia and form 
new amino-acids. Embden and Schmitz were able to demon- 



*® See Lusk: /or. cii. 

" Cathcart: loc. cii., p. 116, ff., (see p. 1, note 1). 



N. W. Janney 343 

strate the synthesis of an amino-acid (alanine) in the perfused 
glycogen-rieh liver, on the addition of ammonium chloride to 
the perfused fluid. Other indirect evidence could be mentioned, 
making it probable that protein may be synthesized from metdboh'c 
products closely related to glucose on the one hand, and simple nitro- 
genmis compounds on the other. 

With the help of this view the apparent urgent need of the starv- 
ing organism for carbohydrate may be explained. Landergren 
found that when the starvation excretion of nitrogen had been 
established, this minimum could be still further depressed bj'' an 
exclusively carbohydrate diet, whereas a diet composed of fats 
employed under the same conditions did not have this effect. 
It is evident from this work that the organism in order to obtain 
carbohydrate even sacrifices an increased amount of its body 
protein. Aside from dynamic questions, this fundamental re- 
quirement of the starving body for glucose may be but an expres- 
sion of its need for prot<5in repair. Certain more vital organs 
are better sustained in starvation than other less important ones. 
It may be that the proteins of the latter are broken down suc- 
cessively to amino-acids, glucose, oxyaldehydes, and ammonia, 
from which the repair requirements of the vital proteins can be 
satisfied by synthesis of the special amino-acids needed. The 
introduction of ingested glucose limits protein breakdown, as it 
represents material suitable to protein synthesis. Osborne and 
Mendel" have successfully maintained animals, which have even 
raised young, on a diet containing a single isolated protein. The 
large number of protein syntheses which must here take place 
can certainly be easily explained by the above view of protein 
metabolism. 

SUMMARY. 

Part I. Vegetable and animal proteins under optimal con- 
ditions are metabolized at the same rate in the animal organism. 
All the extra glucose and nitrogen arc eliminated by the ninth 
hour after ingestion. 

Part II, See italics, first page of this article. 

" Osborne and Mendel : this Journal^ xiii, p. 233, 1912-13. 



344 Metabolic Relationship of Proteins to Glucose 



PROTOCOLS. 



The protocols of the longer periods of these experiments are grouped with the 
succeeding ones. In this series the proteins were fed within the first 
six minutes of the first hour of the partition period. 







S ^K 


8!5 








s 


S 0^ 


oS^ 


y. 




PF.BIODS 


H 2 f* o 

5: 1 c» 1 g« 


OLUC 
HOU] 


• m 






hra. 


gm. gm. \ gm. 


1 

gm. 




8-9 a.m 


< 0.408 


3.390 8.31 


Casein Experiment 1 . At 8 


9-10 a.m 


i , 0-505 


3.950 7.83 


a.m. 25.54 gm. (N-3.88 


10-11 a.m 


0.624 3.623 5.80 

1 i 1 


gm.) casein with 350 cc. 


11 a.m.-12 m. 


0.467 


2.691 


5.76 


water. 


12 m.-l p.m. . 


' 0.442 


2.353' 5.33 




1-2 p.m 


1 0.449 


2.146,' 4.78 




2-3 p.m 


0.378 


1.899i 5.02 




3^ p.m 


' 0.340 


1.508, 4.72 




4-5 p.m 


< 0.336 


1.543 4.59 




5-8 p.m 


0.981i 3.984 0.327 


1.328 4.06 




8-11 p.m 


0.981 4.000' 0.327 


1.333' 4.07 




llp.m.-2a.m. 


1.128 3.922; 0.376 


1.307 3.48 




2-8 a.m 


2.229 7.519, 0.371 
0.^2 


1.253 3.37 
1.117, 4.16 




8-9 a.m 


Casein Experiment 2. At 8 


9-10 a.m 


! 0.408 


1.887 4.63 


a.m. 16 .77 gm. ca8ein(N = 


10-11 a.m 


; 0.496 


2.091' 4.21 


2.56 gm.) with 230 cc. 


11 a.m.-12 m. 


0.467 

■ 1 


1.754 3.76 


water. 


12 m.-l p.m. . 


' 0.408 


1.415 3.47 




1-2 p.m 


0.437 


1.167, 2.73 




2-3 p.m 


, 0.303 


0.935 


3.09 




3-4 p.m 


0.265 0.791! 2.98 




4-5 p.m 


1 0.322 


0.805 2.50 




5-8 p.m 


0.925 2.049, 0.308 
0.743 2.475; 0.248 


0.683 2.22 




8-11 p.m 


0.825, 3.33 
0.666, 2.74 




11 p.m.-12 m. 


0.729 2.000 0.243 

1 1 




2-8 a.m 


1.279 4.167 0.213 


0.695 3.26 




9-10 a.m 


I 0.465 


2.113 4.54 At 9 a.m. 29.29 gm. aerum 


10-11 a.m 


i 0.528 


2.508 4.56 


albumin (N=4.60 gm.) 


11 a.m.-12 m. 


i 0.534 


2.351 4.32 


moistened with a little 


12 m.-l p.m.. 


0.646 


2.164 3.30 


water. 


1-2 p.m 


, 0.729 


2.222 3.00 




2-3 p.m 


0.708 


2.066 


2.92 




3-4 p.m 


0.580 

1 


1.655 


2.85 




4-5 p.m 


, 0.530 


1.500 


2.83 




5-6 p.m 


1 0.450 


1.364 3.03 




6-9 p.m 


1.072 3.540, 0.357 

1 


1.180 3.30 




9 p.m.-12 m. . 


1.080 3.393 0.360 


1.131 


3.14 




12 m.-3 a.m . . 


1.088 3.695 0.362 


1.232 3.40 




3-8 a.m 


1.745 5.334 0.349 


1.066, 3.06 





N. W. Janney 



345 



PROTOCOLS — Concluded. 



PERIODS 
ktt. 

8-9 a.m I 

I 

9-10 a.m 

10-11 a.m 

11 a.m.-12 m. 

12 m-1 p.m..> 

1-2 p.m 

2-3 p.m 

3-4 p.m i 

4-5 p.m 

5-8 p.m ' 

8-11 p.m j 

11 p.m.-2a.m. 

2-5 a.m i 

5-8 a.m 

7-8 a.m 

8-9 a.m 

9-10 a.m 

10-11 a.m 

11 a.m.-i2 m. 

12 m.-l p.m.. 

1-^ p.m 

2-3 p.m 

3-4 p.m 

4-5 p.m 

5-8 p.m 



3 

O 

o 



gm. 



s 

u 

P 
■J 
o 



gm. 



1.521 
1.3951 
1.598' 

1.416; 

1.59li 



4.64^ 
4.279| 
5.000j 
4.828 
4.990 



1.194 2.871 






gm. I 

0.488 

0.513' 

0.746| 

0.828 

0.858 

0.718 

0.660 

0.575 

0.498 

0.507' 

0.465^ 

0.533i 

0.472 

0^1 

0.412 

0.508 

0.81li 

0.863' 

0.653 

0.488 

0.453 

0.410 

0.391 

0.371 

0.398 



3 O 



gm. j 

2.826: 
3.3431 
4.271 
3.127 
3.032 
2.371 
1 .973I 
1.734 
1.750 
1.5481 
1.426J 
1.666 
1.609 
1.663 
"l.225| 
2.230; 
3.509| 
3 . 175| 
1.887 
1.250' 
1.1241 
1.079 
1.183 
1.005 
0.957 



• « 

wtm 



5.79 
6.52 
5.73 
3.77 
3.53 
3.30 
2.99 
3.02 
3.52 
3.05 
3.06 
3.13 
3.41 
3.14 

2.97' 

4.43 

4.33 

3.68 

2.89 

2.56 

2.48 

2.63 

3.02 

2.71 

2.40 



At 8 a.m. 17 .48 gm. gliadin 
(N = 3.07) in 150 cc. fi 
NaOH followed by equal 
volume of /^ HCl. 



At 8 a.m. 16.90 g;m. edestin 
(N=3.18 gm.) with 270 
cc. H2O. 



346 Metabolic Relationship of Proteins to Glucose 



HI'BBTANCE VKD 



NTTROOEN IN 

SUBSTANCK 

FED 



o o 
so 



I 



gm. I gm. ' kgm. 

Casein, 56.8...; f8.88 (l.Ol 8.8 

gm. per 
kgm.) 



Casein, 25.0... 3.90 



6.6 



Casein, 28.4 



Casein, 54.2. 



4.44 + ^ 
0.06 gm. 
beef ex- 
tract N 
= 4.50 . 

8.47 4- " 
0.14 beef 
.1 ^ extract 
N = 
, 8.61 
3.61 4- ] 



7.5 



14.3 



Casein, 23.1 



0.09 beef 
extract 
N = 3.70 



5.7 



9.2 



Casein, 34.7... 



5.28 



9.7 



Casein, •25.54... 3.88 



6.4 I 



Casein,** 16.77.; 



2.m 



8 

o 

a 
J2_ 

kr». 

12 
24 
12 
J2 
24 
24 
24 
24 

12 
24 
24 

12 
12 
24 
24 

12 
24 
24 

24 
24 
24 
24 
12 
12 
12 
12 
12 

12 
12 
12 
12 



I - 



H 

O 

s 

SB 



fa 

8 

o 



z 

6 



gttt. gtn, 

4.30 15.05 3.50 i 

13.08 45.61 3.43 

4.30 13.23 3.07 

3.82 12^93 3.38 

6.87 19.62 2.86 

9.88 27.89 2.82 
8.18 22.44 2.74 
6.64 18.17 2.85 



EXTRA 
OLVCOSE 



gm. 'par^cent 



31.2 54.9 



10.4 41.6 



3.18 
7.86 
5.61 

11.44 
10.31 
24.85 
18.57 
17.78 



11.05 
27.48 
18.41 

I 

I 

35~21 
31.27 
73.14 
58.08 
53.53 



3.38 

3.50 15.89 55.9 

3.28 ! 



3.91 11.87 

. 8.02| 24.23 

6.05| 16.08' 

'~9Tr2'29^74 I 

9.36|31.85 
' 11.0835.80 ' 
_7^87|26^.19 ' 
3.9215.33 
3.9l]l5.73 
4.93|27.09 ' 
4.34;i5.44 
5.32,16^28J 
2.86, 8.66 , 
3.05! 8.20 



4.31 
2.75 



12.645, 
8.64 



3.15 
3.03 
2.94 
3.12 
3.01 

3.04 
3.02 
2.66 

3.26 
3.33 
3.22 
3.33 

3.91 
4.02 
5.49 
3.29 ' 
3.05 ' 

3'06 ; 
2.74 , 
2.93 : 
3.14 i 



23.12 42.6 



12.13' 52.5 



16.49' 47.5 



11.53; 45.1 



7.33. 43.7 



* Casein Partition Experiment 1. In this experiment the nitrogen 
excretion is so irregular that the extra glucose was calculated from the 
augar values alone, which are unusually regular. 

** Casein Partition Experiment 2. 



N. W. Janney 



347 



BVBtrtASGt FKD 



lis 



Ik 

o 
m 






gm. 



ffm» I hQfn> 



Ovalbumin.* 31.70. . .i 4_^2^l 8_^ 

r ■ 



Ovalbiunin, 27.66. . . 4.30 8.6 



Ovalbumin, 18.00... 2.80 I 5.6 



Ovalbumin, 31.42.. . 4.80 i 9.6 



Serum albumin, 34.12 5.20 j 10.4 
Serum albumin, 30.76 4.85 i 9.7 






Serum albumin, 21.93, 3.45 6.9 



Serum albumin, 27.37i 4.28 ' 8.8 

. . I . . . 
Serum 

albumin, •♦29.29... 4.60' 9.2 



8 

o 

a 

E_ 

krt. 

12 
12 
24 
12 
12 
24 
24 

12' 

12 

24 

24 

12~ 

12 

24 

12 

12 

12" 

24 

24 

24 

24^ 

24 

24 

24 



H 
O 
O 



24 
24 
2£ 

12 
12 
12 

12 



gm, I 

3.85j 
3.92 
_8.80' 
3 .55| 
4.05| 
9.44 
8.001 
2.59 
2.54' 
5.80| 
4^37 

4.71 
10.26' 

4.1^ 

3.76 
~6~M 
13.68 
10.98, 
13 . 19 

9.29! 
"8.02' 

9.31| 

7.34 

_i 



pa 

§ 

P 

o 
gm, I 

15.65 
15. 40! 
34.48J 
12.79 
13.6l| 
31.79, 
26J9 
7.71' 
7.51 
18.85 
^1J6; 
19.08 
17.19 
34.55, 
13.90' 
11.88 

23 .'26 
46.87 
36.10 
41.24 
27. 2J 
27.0^ 
32.79 
24.91' 



o 



7.90 
9.57i 
7^62 

4.73 
6.24 
4.17- 
4.04! 



24.971 
31 .75| 
23.53 
17.01| 
21.48' 
I3.31I 
13.89 



4.06 
3.93 
3^2 

3.60 
3.37 
3.37 

2.97 
2.96 
3.25 
2^60 
3.52 
3.65 
3.37 
3.33 
3.16 
3.35 
3.43 
3.29 
3.13 
2^ 

3.39' 

3.52 

3J0 

3.16 

3.32 

3.09 

3.6O' 
3.44 
3.19 
3.43 



EXTBA 
GLUCOSE 



gm. per cent 



17.40 54.9 



14.50 52.14 



10.51 



58.4 



15.95 



50.8 



18.80 
15.30 



55.1 

49.7 



12.87 58.7 



15.25i 55.7 



15.90 53.7 



* In this experiment a partially purified preparation of ovalbiunin (N= 
12.39 per cent) was employed. The other experiments were carried out 
with preparations as described in the text. 

** Partition experiment. 



348 Metabolic Relationship of Proteins to Glucose 



SUB8TAKCE FED 



SB U 

►- o 

Y. < 

PU>H , SO 



o 



9m. 



gtn. I Avm. 



Gelatin, 25.87 1 4.55 9.1 

I 

I 

Gelatin, 38.41 ^ 6.80 13.6 

Gelatin, 22.37 : 3.96 13.2 

i 
I _ 

Gelatin, 27.93 4.95 9.9 



Gelatin, 12.80 2.25 4.5 



Fibrin, 21.40 3.55 ' 7.1 



Fibrin. 19.89 3.30 6.6 



Fibrin, 30.45 5.05 10.1 

Fibrin, 15.06 2.5 5.0 



Edestin, 16.00 3.00' 6.0 



Edestin, 15.47 2.90 



5.8 



Edestin, 26.40 4.95 9.9 



S 

o 
S 

H 

pu 

hrt. 

24 
24 
2A 
24 
24 
24 
24 
24 
24 
24 
24 
12 
12 
12 
24 
24 
24 
24 
24 
24 
24 
24 
24 
24 
24 
24 
'24 
12 
24 
24 
24 
24 
24 

24 
24 
24 
24 



u 

o 

M 

H 

V. 



gtn. 

9.29 

11.26 

7.10 

13 . 14 
13.77 
15.90' 
12.45 

9.26 
12.19 
13.22, 

9.72 

5.36, 

6.23 

4.55 

7.08 

6.28 

7.79 

5.80 

7.16 

4.95 

10.82 

10.43 

12.84 

5.72 
0.76 
4.81 

6.43 
2.98 
7.42 
3.98' 
6.59 
4.31' 
4.31 

10 . 19 

11.57 

S.6O' 

8.03 



S 

•J 



27.21' 
36.70 
19.42' 
44^451 
45.98 
56.43 
43.32 
35.81 
41.24 
47.34 
34.07 
17.75 
20.30 
12.80 
20.36 
18.26 
24.69 
17.83 
23.46| 
19.37 
35.09. 
36.47 
42.85 
21 .56 
22.99 
16.04 
25.56 
12.65 
27.70 
16.67 
26.22 
18.45 
16.52 

36.53! 
41.33 
34.62 
29.23 



• • 

o 

2.93 
3.25 
2.72 
3.38 
3.34 
3.54 
3.48 
3.87 
3.38 
3.58 
3.51 
3.31 
3.26 
2.82 

2.88 
2.91 
3.17 
3.07 
3.28 
3.91 
3.24 
3.50 
3.34 

3.76 
3.40 
3.33 
3.97 
4.25 
3.73 
4.18 
3.96 
4.28 
3.83 

3.58 
3.57 
4.02 
3.64 



I 



EXTRA 
QLUCOeE 



gm. percent 

I 

17.78. 68.7 



24.40^ 63.5 
13.44! 60.1 



1 



18.85' 67.5 



8.12 63.4 



12.01, 56.1 
10.00 50.3 



16.59 54.5 
7.91' 52.5 



9.54 50 6 
10. 6I; 6S 6 



7- 17.30 65.5 



N. W. Janriey 



349 



•UBflTANCB n>D 






H 

e 

H 



fifl. 



Edestin, 14.56. 



ftn. 



KQilt, 



2.73 5.5 



Edestln, 16.90 



3.18 7.9^ 



Gliadin,* 19.66 < 3.36 



11.2 



GUadin,* 16.45 2.94 ; 9.8 



Gliadin, 22.6 i 4.10 8.2 



Gliadin, 11.70 ,2.07 ! 6.9 



GUadin,** 17.48 '3.07' 8.3 



8 

o 
S 



hr§. 

24 
24 
24 
12 
12 
J2 
12 
24 
24 
24 
2l4" 
24 
24 
12 
12 
24 
24 



Zein,*** 20.68 3.36 10.3 



12 
12 
12 
12 

12 
12 
24 
24 



as 

H 

O 

s 



pa 

! 

p 

■J 
o 



gm. I 

6.94i 
8.13] 
5^72i 
3.89^ 
6.5^ 
4.15; 

3.68 
7.11 
7.90 
6^63 

9.73 

10. 2d 

_7^76' 

4.'2(> 

4.18. 

10.28 



o 



gm, ] 

24.94 
29.52! 
21^56 
11.56 
20.54' 
11.46! 
13.18' 
27.08| 
32.00| 
22.22' 



BZTBA 

QLUCOeB 



12 


4.61 


12 


5.45 


12 


4 05 



6.84 
7.49 
6.001 
6.21 



5.12| 
4.65! 
9.27I 
8.7oi 



30.81 
36.57j 
24jr5 
13.96, 
14.69 
38.04 
26.69 
11.74 
17.32i 
9.99' 
22.36 
29.07 
19.10 
19^! 
15.70, 
15.54' 
31.09; 
28.44' 





gm. 


3.60 




3.63 


9.65 


3.76 




2.97 




3.14 


10.91 


2.76 




3 58 




3.81 




4.05 


16.75 


3.35 
3.17 





3.59 


13.48 


3.19 





3.32 
3.52 
3.70 
3.15 

2.55 
3.17 
2.47 



3.27 
3.88 
3.18 
3.14 

3.45 
3.35 
3.35 
3.25 



17.77 



ptr cent 

66.3 
64.6 



80.1 



82.0 



78.6 



8.84 75.6 



14.83 84.8 



11.47: 55.5 



* In these experiments gliadin was fed as moist boli on the back of the 
tongue. In the others it was dissolved in acid or alkali (see text). From 
30 to 66 cc. of total fluid per kgm. were employed. 

*• Partition experiment. 

*** In this experiment the granulated zein was fed as moist boli on the 
back of the tongue. In the other experiments it was freshly precipi- 
tated, as described in the text. 



THBJOUBlfALOPBIOLOaiCALCHKMmTBT, VOL. XX. NO. 3 



350 Metabolic Relatitfnship of Proteins to Glucose 



HL-B8TA.NCE FEU 

- - -- - - 


NITBOOXN IN 
SUBSTANCE 
FED 


WEIGHT OF 
DOG 


1 

1 PERIODS 


1 
1 H K 1 

S 5 ' o , 


E.XTRA 
GI.UCOSi: 




'jtn. 


gm. 


KQin. 


hrt. 

12 
12 


gm. gm. 

3.98 12.58 3.16 
3.81, 11.43, 3.00 


gm. per cent 


Zcin, 


17.23 


2.80 , 


5.2 


24 
12 


7.02 21.28, 3.03 

3.03 9.05 2.95 


8.64 50.2 






x^ • \^ ^ ^^^^ • ^ 








1 


12 


3.00 8.10,' 2.70 , 








i 


1 

1 


12 
12 


3.15 9.05 2.87 
2.89 7.57 2.67 




Zciii, 


14.68 1 


2.39 1 

1 


4.9 1 


24 
12 


5.84 16.95 2.90 
1.95 5.66 2.90 


7.26 49.4 








Zcin, 


7.93 


1.29 

• 


4.4 


24 
12 
12 
12 
12 


. 4.23 12.18 3.03 
1.88 4.59 2.44 
1.75 4.33 2.47 
5.81 17^54 3.02 
5.47 16.19 2.96 


4.39 55.3 


Zoin, 


16.55 


2.70 ; 

1 


9.2 


24 
12 
1? 


11.44 34.68; 3.03 
5.48 16. 27^ 2.97 
4.61 13.16' 2.85 


8.90 53.8 


Zcin, 


15.08 


2.46 


8.4 


12 i 
12 


5.33 16.13 3.03 
4.37 11.56 2.64 


8.15 54.0 
















12 


4.16 11.05 2.65 












12 


3.80 10.65 2.92 





TBDB COMPARATIVE NUTRITIVE VALUE OF CERTAIN 

PROTEINS IN GROWTH, AND THE PROBLEM 

OF THE PROTEIN MINIMUM.^ 

By THOMAS B. OSBORNE and LAFAYETTE B. MENDEL. 

With the Co&peration of Edna L. Ferrt and Alfred J. Wakeman. 

{From the Laboratory of the Connecticut AffrictUtural Experiment Statioli 
and the Sheffield Laboratory of Physiological Chemistry in 

Yale University, New Haven.) 

(Received for publication, January 30, 1915.) 

In earlier communications^ we have pointed out the dominant 
importance of certain amino-acids in the problems relating to the 
function of the nitrogenous food intake in both maintenance and 
growth. When the content of any essential amino-acid group in 
a specific protein is relatively small, the comparative poverty in 
the amino-acid in question will not manifest itself so long as the 
diet contains a surplus of this amino-acid above either the 
maintenance or growth quota. If, on the contrary, the intake 
of the protein is kept law, a plane will ultimately be reached where 
the yield of the amino-acid in question becomes so small that it 
cannot satisfy, first, the growth requirement and, later, the main- 
tenance need for the nutrient unit in question, even though the 
other amino-acids are still available in suitable quantity. With a 
coDstaoit energy intake the amount of protein available for con- 
structive functions will be limited by the "law of minimum." 
For example, a diet containing 20 per cent of the calories ingested 
in the form of some protein relatively deficient in an essential 
amino-acid may supply enough of that amino-acid to satisfy the 
requirements of the animal for maintenance and growth; whereas 

* The expenses of this investigation were shared by the Connecticut 
Agricultural Experiment Station and the Carnegie Institution of Wash- 
ington, D. C. 

* T, B. Osborne and L. B. Mendel: this Journaly xvii, p. 325, 1914; 
xviii. p. 1, 1914. 

351 



352 Nutritive Value of Proteins in Growth 

one containing only 10 per cent of its calories in the form of the 
same protein, with an equivalent energy intake, may not supply 
enough. To determine whether or not such deficiencies will be 
shown by experiments in which rats are supplied ad libiUnn with 
foods contaming different percentages of protein, but of approxi- 
mately the same calorific value, we have tried experiments the 
results of which are shown in the following charts in the appendix: 
I, with 18, 15, 12, 9, 6, 4.5, and 2 per cent of casein; II, with 18, 
15, 12, 9, 4.5, and 2 per cent of edestin; and III, with 18, 11, 9, 
4.5, 2.5, 2, and 1 per cent of lactaJbumin, 

To learn the cause of the slower growth on the diets containing 
f^ the lower percentages of protein we have tried the effect of addi- 
tions of cystine to the casein foods and of lysine to the edestin 
foods, with the results shown in Charts IV and V. On compari- 
son with many other proteins casein doubtless yields relatively 
little cystine; for on treating with caustic alkali according to the 
method of Schulz* only one-eighth of the total sulphur is con- 
verted into sulphide. This proportion is much less than that 
jrielded by most other proteins.* Thus, of the proteins used in 
our experiments, gliadin and conglutin yield about two-thirds, 
glycinin and edestin about one-half, zein, ovovitellin, and ov- 
albumin about one-third of their total sulphur as sulphide. 
Since, under the conditions existing in the analytical processes 
to which the proteins were subjected in making these determina- 
tions, cystine yields about two-thirds of its sulphur as sulphide, 
it is not improbable that from these figures the relative pro- 
portion of the cystine complex in these proteins can be calculated. 
Assuming that the sulphur which separates as sulphide cor- 
responds to two-thirds of the sulphur of the cystine complex which 
they contain, casein should yield about 0.60 per cent of cystine, 
whereas ovalbumin should yield 3.0 per cent and gliadin 3.7 per 
cent. Attempts to isolate cystine directly from the proteins 
have thus far yielded only a small part of the total cystine which 
they probably should. 

In order to determine the effect of adding an amino-acid to a 
food containing a protein yielding this amino-acid in a relatively 
small amount, it is necessary to know the least quantity of this 

' N. Schulz: ZtHchr.f. physioL Chem.y xxv, p. 16, 1808. 

* T. R. Osborne: Jour, Am. Chem. Soc, xxiv, p. 140, 1902. 



T. B. Osborne and L. B. Mendel 



353 



protein which will promote normal growth. Obviously if more 
is supplied the additions will have no eflFect on the rate of growth. 
Furthermote, definite conclusions can only be reached when the 
comparisons arc made during periods in which the amount of 
food eaten is the same. In the experiments here described these 
conditions have only been approximately fulfilled, and it will be 
necessary to adopt a different procedure before satisfactory evi- 
dence can be obtained which will demonstrate in detail that which 
in general these experiments unquestionably show. 

In regard to the minimum amount of casein which the food must 
contain in order to promote normal growth, we have established 
by numerous experiments that when the diet contains 18 per 
cent of casein, along with the essential non-protein components 
of the ration, young rats can complete their growth satisfactorily. 
We have brought such anhnals to full size and kept them in health 
on such a diet for more than 620 days. With 15 per cent of casein 
in the food growth was still made at a normal rate (see Rat 2465, 
Chart I), although the summary below shows that a smaller 
amount of casein was eaten : 



GREW 



ATB 



BLAT 



POOD 



1592 
2465 



18% casein 
15% 



( t 







OAJN 




from 


to 




Food 


gm. 


gm. 


gm. dif$. 


gm. 


83 


287 


204 


119 


1228 


77 


286 


209 


119 


1187 



gm. 

199 
160 



When the casein content is reduced to 12 per cent growth falls 
a little below the nonnal, as shown by Rat 2117, Chart IV. The 
failure of Rat 2117 to grow as well on the 12 per cent casein food 
as did Rat 2465 on the 15 per cent casein food might have been 
due to a too low food intake; or conversely the lower food intake 
might be attributable to the failure to grow and consequent need 
for less protein. That the former is not true is shown by the 
normal growth of Rat 2124, Chart IV, which had the same food 
as Rat 2117, but with an addition of cystine equal to 3 per cent 
of the casein. 



354 Nutritive Value of Proteins in Growth 





j 


GREW I 






1 ATE 


RAT 


FOOD 




OAIN 








from to 

i__ 




, ^ood Cojtoin 




; 


gtn. gtn. . 


gm. 


iy*. 


1 ^ ^ . 

gvi. gtn. 


2465 


15% casein 


103 . 208 


105 


56 


498 1 67.0 


211 r' 


12% " 


98 ' 142 1 


44 


56 


, 277 ! 40.6 


2124 


12% " 




1 


i 


1 




1 -f cystine 


101 


181 ' 


80 56 

1 


; 388 41.1 




_ ._. 


_ .. _ 











Since the addition of cystine in this case rendered the 12 per 
cent casein food more efficient for growth, it is probable that the 
minimum proportion of casein in food mixtures of this character 
for normal growth lies somewhat below. 15 per cent. In how far 
differences in the ability of individual rats to utilize their food for 
growth may modify such a conclusion can only be determined by 
further experiments. 

When the casein is reduced to 9 per cent, growth is promptly 
limited by the protein factor. Our present studies* have made it 
obvious that this failure to grow at a rate^equivalent to the hormal 
is not attributable to too little protein in the diet. The addition 
of isolated cystine to the food containing 9 per cent of casein, without 
any other supplement, at once renders the ration decidedly more 
adequate for growth. This is well shown by comparison of Rats 
2043, 2051, 2481, 2483, 2484 in Chart IV. Growth can be faciU- 
tated or repressed at will by the addition or withdrawal of the 
extra cystine from the diet containing 9 per cent of casein. 

Obviously a progressively lowered intake of casein with its 
detrimental consequences for growth cannot be remedied inde- 
finitely by supplementing it with cystine. A level is finally 
reached at which the possibility of further growth, or even of 
maintenance, is limited by the lack of many, or all, of the numer- 
ous necessary nitrogenous units. The protein intake as a whole 

' The data recorded here are not directly comparable with those relat- 
ing to limiting growth by diminishing the proportion of protein which 
were published earlier by us (Zischr. f. physiol. Chem., Ixxx, p. 340, 1912), 
because the foods in the present series contained butter-fat, which has 
been shown to facilitate growth (this Journal, xvi, p. 423, 1913-14). In 
the former experiments lard was the sole fat used, and failure to grow in- 
variably ensued sooner or later on these diets, quite independently of the 
content of protein present. 



T. B. Osborne and L. B. Mendel 



355 



becomes inadequate. Other amino-acids are simultaneously 
needed. It will be noted in Chart IV that when the food con- 
tains only 6 per cent or 4.5 per cent of casein (Rats 2519, 2116), 
added cystine alone no longer suffices to facilitate growth as 
vigorously as it did in the case where 9 per cent of casein was 
present (Chart IV, Rats 2043, 2051, 2481, 2483, 2484). 

Comparable phenomena were obtained when edestin formed 
the sole protein of the diet (see Chart V). Here, too, growth on 
a ration containing only 9 per cent of edestin has fallen behind 
that secured with larger proportions. That the limiting factor 
may be the comparatively low yield of lysine is indicated in the 
figures below:® 

Lactalbumin, cow's milk 8 .10 per cent 

Casein, " " 7 .61 per cent 

Edestin, hempHseed 1 .65 per cent 

Gliadin, wheat 0.16 per cent 

As with casein, normal gi'owth has been secured with foods 
containing 15 per cent of edestin. With 12 per cent the rate of 
growth was a little less than normal. That this proportion of 
edestin is somewhat too small is indicated by the slightly improved 
growth when lysine equal to 2 per cent of the edestin was added 
to the food. That this was not attributable to an increased food 
intake is seen from the following figures: 









GREW 


1 

j 








1 

ATE 


RAT 


LYSINE 


1 




i 

»o 1 




GAIN 




, Food 






from 


1 

1 


Edestin 






gm. 




gm. 


gm. 




dya. 


j gm. 


gm. 


2176 





53 




129 


76 




70 


435 


48.5 


2120 





54 




121 


67 




70 


391 


43.6 


2119 


+ 


51 


1 


144 


93 


1 


70 


389 


43.4 



From these data it would appear that the added lysine renders 
the edestin in the food containing 12 per cent slightly more 
available for growth, and that this proportion of edestin is some- 
what too small to meet the full requirements of these animals, 
for none of them grew at the full normal rate (see Chart V). 



* A more extensive tabular summary is given in our paper: this Journal 
xvii, p. 334, 1914. 



356 Nutritive Value of Proteins in Growth 

* With 9 per cent edestin, as can be seen from Chart V, growth 
was less rapid than with 12 per cent. The addition of lysine to 
the 9 per cent edestin food caused some improvement in growth. 
This increased growth was relatively much greater than can be 
attributed to the increased food intake, especially when the greater 
requirement for food caused by increase in size of the growing rat 
is taken into consideration. Thus: 



t 




ORBW 










▲TB 


BAT 


LTSINB 






GAIN 

gm. 




Food 






from 


to 


Edestin 

1 




gm. ' 


gm. 


iv. 


gm. 


1 

gm. 


2110 





59 


84 


25 ; 


42 


239 


; 20.0 


2050 





48 1 


81 


33 


42 


234 


; 19.6 


2110 





93 


126 


33 


42 


264 


22.1 


2050 


+ 


91 


172 


81 


42 


332 


i 27.8 






• _ 




1 









Comparing 2110 with 2050 in the second period, the latter made 
a gain in weight 145 per cent greater than that made by 2110, 
but ate only 22 per cent more food, despite its larger ultimate size. 

The favorable results following the addition of lysine to the 
9 per cent edestin food are not as striking as in the case of the 
cystine-casein experiment. It is, of course, not difl&cult to con- 
jecture that the relative paucity of lysine groups in edestin is not 
as marked as is the comparatively great deficiency of the cystine 
in casein. In either case it is not more protein as a whole, i.e., 
not all of the amino-acids that are required to make the con- 
structive material adequate; though of course the deficiency can 
be supplied as well by raising the proportion of the protein it- 
self as by additions of amino-acids. 

The relatively greater efficiency of Uwtalbumin in promoting 
grou^th is striking; for with only 9 per cent in the food the rate 
of growth was about normal (Chart VI), and the lower percent- 
ages were in all cases far more efficient than the corresponding 
proportions of casein or edestin (Charts VII. and VIII). Owing 
to our present limited knowledge of the products of hydrolysis 
of lactalbumin no experiments were made with amino-acid 
additions. Such information as we have relating to its amino- 



T. B. Osborne and L. B. Mendel 357 

acid make-up shows that both lysine and tryptophane^ are rel- 
atively abundant. The marked nutritive efl&ciency of lactal- 
bumin is probably due to a more perfect balance in the pro- 
portions of the essential amino-acid groups which it contains. 

Relation of growth to food intake. It may be objected fairly 
that some of the conclusions here drawn are rendered inconclusive 
by the uncertain factor of the total food intake of the rats. Obvi- 
ously if an animal consumes twice as much of a 5 per cent protein 
food as of a comparable 10 per cent protein ration it will obtain 
precisely the same total amount of protein daily from these food 
mixtures of quite unlike percentage composition. To permit 
tenable conclusions it is necessary to ascertain the actual intake 
and calculate the absolute amounts of protein ingested. Rat 
2051, Chart IV, for example, on a diet containing 9 per cent of 
casein gained 8 grams in body weight in a period of three weeks, 
with a food intake of 113 grams (9.2 grams of protein); on the 
diet containing 9 per cent of casein plus cystine the same rat 
gained 35 grams of body weight in a second period of three weeks 
on a food intake of 123 grams (10 grams of protein). In other 
words, the growth in the presence of cystine was nearly 400 per 
cent greater, — a gain which could hardly be accounted for by the 

8 per cent increase in the amount of food consumed, and despite 
the need of more food on the part of the animal as it became 
larger in the second period. Again, Rat 2481 on food containing 

9 per cent casein plus cystine, in 35 days ate 317 grams of food 
(containing 25.4 grams of protein) to grow from 97 to 164 grams, 
I.e., to gain 67 grams in body weight; whereas Rat 2117 (Chart 
IV), on 12 per cent casein food in 84 days ate 584 grams (con- 
taining 63.2 grams of protein) to make essentially the same gain, 
I.e., to grow from 98 to 162 grams. Expressed in stili another 
way, during the five weeks in which Rat 2481 made the 67 
grams of body increment on a cystine food intake of 317 grams, 
Rat 2117 gained only 38 grams on a food intake of 247 grams 
containing essentially the same amount of protein, viz.y 27 grams. 
A quantity of food of approximately the same order of magnitude 

' Unpublished determinations of tryptophane from lactalbumin made 
in this laboratory have shown higher yields of this amino-acid than have 
been obtained from any other protein as yet thus examined. 



358 



Nutritive Value of Proteins in Growth 



from the standpoint of its energy yield has produced better 
growth in various instances, summarized below: 



1655 d* 
1654d^ 
1615 d^ 
1657cf 
1592 d^ 
1650 d^ 



RAT 



IXITIA.L 
WEIQHT 

gm. 

91 
91 
93 

104 
98 

100 



WEIOHT 
GAINED 

gm. 

58 
65 
50 
73 
80 
73 



FOOD CATSN 

gm. 

237 
251 
253 
284 
285 
285 



DURATION OF 
OBSERVATION 

dps. 

35 
35 
35 
35 
35 
35 



It is not always easy to dissociate the increased growth on a 
particular diet from the factor of increased food intake thereon; 
for when the animal actually increases in body weight it requires 
more food, and the comparison of the larger animal on the suit- 
able diet with the smaller one on an unsuitable ration is almost 
impracticable. We have the impression that the effect of cystine 
is something more than a mere stimulus to appetite leading to 
greater food intake. This belief is fortified by the observation 
that the rats which grow well on 9 per cent casein plus cystine in 
the food frequently eat no morp in terms of calories and ingest no 
more protein than do rats which are growing no better on even 
larger percentages of casein without the addition of cystine. It 
thus appears that a marked deficiency in any essential ingredient 
of the diet does not lead to a corresponding compensatory increase 
in food intake.* 

The experience which we have gained in measuring the food 
eaten by many rats at all periods of their growth has given us 
the conviction that the intake of the individual is determined in 
large measure by the energy requirement at any given period. 

* This fact may help to explain the nutritive failures recently reported 
by P. Tachau {Biochem, Ztschr., Ixv, p. 253, 1914) for mice fed on rations 
in which a suitable diet was made unsuitable by extensive addition of 
non-protein components, — fats and carbohydrates. To us, the consequent 
lowering of the per cent of protein in the ration without a corresponding 
increment in food consumption to keep the protein up to the requisite 
minimum seems a likely explanation of Tachau's results. This is forti- 
fied by the fact that additions of protein to the unsuitable food produced 
improvement. 



T. B. Osborne and L. B. Mendel 



359 



Aside from occasional extreme figures for food intake furnished 
by our records, this seems to be substantially true. 

Food intake and gains in body weight of rats growing on diets containing 

varioiLS percentages of cnsein. 











IN'ITIAL 






IN'TAKE OF 


KA.T 




DIET 


* 


BODY 


«.\IX 


- 












WEIGHT 






Total 
food 


Protein 










gm. 


1 


dys. 


gm. 


gm. 


1655 cf ' 


18% 


casein 




60 


135 


80 


578 


93.7 


1652d'. ... 


it 


t < 




60 


141 1 


80 


667 


108.1 


1592d^ 


(I 


(I 




49 


171 


80 


616 


99.7 


2465 d' 


15% 


casein 




75 


151 ; 


80 


697 


94.1 


24660^ 


H 


(( 




81 


83 : 


80 


595 


80.3 


2117d' 


12% 


casein 




45 


95 ; 


80 


510 


55.1 


2481 cf ; 


9% 


casein 


+ cystine 


51 

1 


135 


80 


667 


54.0 


2483cf 


<( 


(( 


it 


52 ! 


107 ' 


80 


591 


47.9 


2484 d^ 


(( 


t< 


tt 


00 


101 


80 


619 


50.1 


2051d^ ' 


9% 


casein 




42 


59 


80 


435 


35.2 


2519c^ 


6% 


casein 


-f cystine 


52 


65 


80 


511 


27.6 


2509 d' 


6%. 


casein 




53 


31 ! 


80 : 


440 


23.8 


2506 cf 


li 


(( 




53 


15 


80 


375 


20.2 










' 


i 









These figui*es show the actual food intake of a series of growing 
rats at different stages of growth. They give some idea of the 
extent to which the preceding statement is valid. At any rate 
the animals do not consume proportionately more of the artificial 
ration because it happens to be decidedly poor in protein; but, 
roughly speaking, they apparently limit their feeding to the amount 
of food jrielding approximately the requisite energy. They may 
even eat less of it than of the ration richer in protein. If we com- 
pare the intake of food containing 18 per cent or 15 per cent of 
protein with that of food containing only 9 per cent, but with 
addition of cystine, it is evident that m order to furnish the same 
quantities of protein in each case it would be necessary for the 
animal to eat respectively two or one and two-thirds times the 
amount of the low protein diet as of the rations higher in protein. 
This has not been the case, as the actual data above regarding 
food intake will show. 

In harmony with the foregoing comments Hopkins has remarked: 
"Only those perhaps who have had the experience of feeding ani- 



360 Nutritive Value of Proteins in Growth 

mals with excess of food, and have noted the amount eaten for con- 
siderable periods, will realize how well adjusted, under normal 
circumstj^nces, is the instinctive appetite to the physiological 
needs. "^ Rubner^® states that many years of experience with 
dogs leads him to believe that appetite and capacity for digestion 
and absorption depend on the dog's requirement for energy in 
his given state of nutrition. A diet which a dog will greedily 
devour when in a room at a temperature of 0°, he will in part 
refuse when at a temperature of 33°. 

A study of the intake of food containing only 2 per cent of pro- 
tein by the rats whose weight curves are plotted in Chart VIII 
shows that in terms of energy it is not widely different, for ani- 
mals of a given size, from the intake of similar food, rich in protein , 
by animals of the same size which are growing satisfactorily. 
A few data may suffice for illustration: 

Food intake of rats on foods of varying protein concentration.^^ 

{Grams eaten per week.) 



RAT 


BODY : 

WKir.HT 






NATURE OF FOOD 




gin. 


















Rats 


groicing. 


16506^ 


85 


18% 


cdestin 


-f- protein-free mi 


1652 d^ 


80 


18':i 


casein 






1654c^ 


83 1 


mi- 


edestin 






1655 cT 


79 


mi 


casein 






15920= 


83 


18% 


n 






1599 a' 


90 . 


18% 


< t 






16199 


■■ 83 ' 


18% 


I, 






16369 


87 ; 


18% 


a 







VOOD 
INTAKB 

gm. 



er-fat 41 

41 
38 
38 
42 
46 
43 
45 



9 F. G. Hopkins: Jour. Physiol., xliv, p. 442, 1912. 

*° M. lluhner: Die Gcnetze des Energievcrbrauchsi hci ErnUhrungy Ijcipsic, 
1902, p. 83, quoted by G. Lusk: The Elements of the Science of Nulriiionf 
2d edition, Philadelphia, 1909, p. 218. 

" The figures are given in gm. rather than estimated calories, because 
the mixtures are essentially alike in their physiological fuel value. Pro- 
tein is replaced by starch or sugar having approximately equivalent energy 
values. The "protein-free milk" contains additional protein equal to 
0.6 per cent of tlie food. 



T. B. Osborne and L. B. Mendel 



S^i 



Food intake of raU — Concluded. 



HAT 



BODY 
WEIGHT 

gm. 



NATUBE OP rOOD 



rooD 

INTAKE 

gm. 



Rata Jailing to grow. 



21049 
2112 9 
21859 
2202 d^ 
2292 cf 
24379 
24509 
24299 
24629 
24309 
24469 
24329 
24619 
24359 
24639 
24289 
24459 



88 
88 
80 
80 
76 
79 
82 
78 
80 
81 
81 
83 
79 
80 
85 
75 
78 



(( 



2% lactalbumin + protein-free milk + 
2% 

2% 
2% 



butter 



-fat 



tt 



^0 

2% 
2^0 

29i 
2% 
2% 
2% 
2% 
2% 
2% 

No 



casein 

n 

edestin 
tt 

glutenin 
i» 

glycinin 

ii 

gliadin 



< ; 



n 

n 

• 4 



t( 



II 



protein 









it 



52 
47 
33 
36 
55 
41 
47 
40 
42 
41 
38 
39 
40 
29 
45 
38 
33 



Comparative efficiency of proteins for growth and maintenance. 
By the methods employed it is possible to compare the efficiency 
of different proteins in promoting growth or maintenance. With 
an abundance of protein in the diet, along with suitable non- 
protein adjuvants, adequate growth has been observed with the 
most diverse proteins, such as: 



Proteint of 
animal origin 

Casein (milk)"- "• "• "• "• i^* »• 
Lactalbumin (milk)»=' »• "• i« 
Ovalbumin (hen's egR)"' *•• " 
Ovovitellin (hen's egg)"- *• 



Proteins of 
tegetabU origin 

Edestin (hemp-seed)"* »• »• »»• "• »^* »« 
Globulin (squashnaeed)"' " 
Excelsin (Brazil nut)" 
Glutelin (maize)"* "• »' 
Globulin (cottonseed)" 
Glutenin (wheat)'^. i3. ii. le 
Glycinin (soy bean)"- " 
Glutelin (hemp-socd)" 



"Osborne and Mendel: Feeding Experiments with Isolated Food- 
Substances, Cnrnvgie Institution of Washinifdnt, Puhlicniion Xo. 166, pt. 
ii, 1911. . 

'See footnotes on pnK<' 3r>2.) 



362 Nutritive Value of Proteins in Growth 

It is interesting to contrast not only the comparative rate of 
growth with different concentrations of the same protein (see 
Chart I for casein, Chart II for edestin, Chart III for lactalbumln), 
but also the unlike ability to grow on diets containing the same 
percentage of different proteins (Charts YL and VII). Criticisms 
which may be applied to this method of comparison will be dis- 
cussed later. 

Closely related to these features of nutrition are the compara- 
tive results obtained with the same concentration of different 
proteins when the nitrogenous intake is kept below the level at 
which adequate maintenance is possible. Even here the gradual 
fall of body weight is unlike with the different proteins (see 
Chart VIII). For further comparison Rats 2428 and 2445 were 
fed on diets otherwise similar, but containing only the exception- 
ally small amount of protein (0.63 per cent) present in the "pro- 
tein-free milk. '' 

The problem of the protein viinivium. In studying the much 
debated question of the protein minimum animals have usually 
been fed with varying quantities of the mixed proteins char- 
acteristic of familiar food products. The foregoing discussion 
of our results with diets containing a single protem naturally 
suggests that the inequalities of the albuminous compounds in 
respect to their amino-acid make-up will give them quite unlike 
values when the minimum quantities for maintenance or for growth 
are approached, the energy intake remaining the same. To cite 
an extreme case, no amount of zein food, however large, will 
enable rats to maintain their nutritive equilibrium. A small 
addition of tryptophane will at once convert the inefficient food 
into a maintenance ration. A rat has been kept without change 
of weight for more than six months on a diet containing zein and 
tryptophane as the sources of nitrogen (see Chart IX) .*° If 

^* Osborne and Mendel: Ztachr. f. physiol. Chem., Ixxx, p. 307, 1912. . 
** Osborne and Mendel: this Jovrnal, xii, p. 81, 1912. 
" Osborne and Mendel: ibid,, xii, p. 473, 1912. 
^* Osborne and Mendel: ibid.^ xv, p. 311, 1913. 
" Osborne and Mendel: ibid.j xvi, p. 423, 1913-14. 
'* Osborne and Mendel: ibid., xvii, p. 401, 1914. 
*^ Osborne and Mendel: ibid., xviii, p. 1, 1914. 

-° For other data of this sort see Osborne and Mendel: ibid.f xvii, p. 
325, 1914. 



T. B. Osborne and L. B. Mendel 363 

a single amino-acid, tryptophane, can play a rdle thus important 
in maintenance, it is more than likely that as the lowest limits of 
requirement are approached the inequalities in the proteins will 
make themselves noticeable in the unlike quantities needed for 
the different physiological performances. For example, with 
the other essential amino-acids equally well provided, the required 
minimum of the protein lowest in its jrield of tryptophane may be 
expected to be greater than that of a protein comparatively rich 
in tryptophane precursors. 

The inequalities of different sources of protein in meeting the 
nutritive needs have been recognized in recent years by various 
investigators. In considering the comparative nutritive values 
of different isolated proteins one is at once confronted with the 
difficulty of comparing the nitrogenous needs of different animals 
of different sizes and sexes at different ages. It is generally 
believed that in growth a liberal supply of protein is required for 
constructive purposes beside what the wear-and-tear functions 
call for. For the present we need not consider the added uses of 
protein as a physiological fuel material. How is the protein 
requirement of animals of unlike size to be measured? What 
unit shall serve as a basis for comparison? Evidently an older 
adult with an abundance of reserve fat cannot be contrasted gram 
per gram of body weight with a poorly nourished adolescent animal. 

In the experiments in this direction which we have thus far 
conducted the animals have had access to unlimited quantities 
of the mixtures of isolated food substances under investigation. 
The amount of food eaten, i.e., the total energy intake, has now 
been ascertained in a large number of instances of normal growth 
on such rations. The word "normaF' is here used as sjoionymous 
with the average rate or curve of growth exhibited by the same 
species and sex living on unlimited quantities of a suitable mixed 
diet. To enable the reader to appreciate the range of variations 
in this matter the grams of food eaten per week by rats which 
were making normal growth have been ascertained at different 
stages of growth on the comparable diets given in the tables on 
pages 365 and 366. In each group the foods of the individual ani- 
mals differ in no respect except with regard to the protein, which is 
casein in some trials and edestin ia others. 

The averages are expressed in graphic form in Chart X of the 



364 Nutritive Value of Proteins in Growth 

appendix. The early consistently smaller rate of food intake of 
normally growing rats on diets containing butter-fat, of which 
the remarkable potency in facilitating growth has been dis- 
cussed elsewhere,*' is perhaps to be expected; for if a food is de- 
ficient in some essential element, it seems reasonable to suppose 
that the animal will endeavor to remedy the deficiency by in- 
creasing its food intake up to the point where excess of food above 
the calorific and other requirements of the animal will bring 
about the well known disturbances of digestion caused by over- 
eating, and consequently put an end to further increase of food 
intake. 

The range of variation in the tabulated results on these strictly 
comparable diets is not inconsiderable; yet when one considers 
differences in the muscular activity of the different individuals, 
they may after all not exceed what this variable factor would 
account for. At present this must remain mere conjecture. The 
ideal method of ascertaining the protein minimum would con- 
sist in feeding exactly equivalent amounts of energ>' in the form 
of foods with unlike proportions of the individual proteins, thereby 
learning from the failure of proper gains where the minimum 
for growth lies, or what the limit of intake for maintenance con- 
sists in, when that is the function under consideration. Until 
recently it has been impracticable, if not impossible, for us to 
follow this procedure, although we now have experiments relat- 
ing to this problem in progress. 

In our observations on the rate of growth as well as the effi- 
ciency of maintenance when rats were fed on food mixtures con- 
taining a different content of protein, ranging from 18 per cent 
(which sufl&ced for adequate growth) to 2 per cent (which has not 
sufficed for maintenance even) (Charts I, II, III, VI, VII, VIII), 
the rate of food intake shows that the failures cannot usually b(* 
explained by a lack of energ^^ in the diet . Frequently rats which 
failed to grow normally were eating sufficient food to enab'e them 
to grow if the protein factor had been adequate.^ In many 
cases the food intake exceeded the average of those given in the* 

** Osborne and Mendel: ibid., xvi, p. 423, 1913; xvii, p. 401, 1914. 
E. V. McCollum and M. Davis: ibid., xv, p. 167. 1913. 

-- For a discussion of the relation of growth to food intake, compare 
Hopkins: Jovr. Physiol., xliv, p. 425, 1912. 



T. B. Osborne and L. B. Mendel 



3^5 



I 









i 

O 



I 



K 

O 



3 

PS 

i 



CO 



CO ic C^ o 
t^ t^ t^ CO 



CO ' 






§ 



sssssss-sss 



s 



s 






CO 



r:iririS^»^^"^^cocooooot^cococo 
oooooooot^t^t^t^t^t^t^r^r^cocoooco 



t^r»r»i^t^t^t^t^t^t^b-t>*b-coS<ococo»<o 



S 



t^t^t^t^t^b-t^t^t^t^t^b-l^<0<0C0C0<0CD<0C03>»0»C 



t^t^t^r«t^b*t^co<ocoococoocococococo^LOt?>i?>u5 



to 






CO 









r*t^t^t^t^t^<oS»co<o<ococodcococococo»otoiLOioio 



eo 



t^cocooSco<o<ocodcococo<ocococoiL!5iotoiou5io 



CO . 



« i 

CO • 



o>t^KdkO'^coi-«i-ii-iQC)Oa>a»OQQt^cocoup-^cocoo>co 

COCOCOCOcOCOCOCOCOOkOAkOiOtOiOaOtOtOtOkOaOuS^^ 





n 


•^ 


*> 1 5? 


*>« 


**-» ft 




00 


o 




•*•» 


n 




o 






>« 




S 






v^ 



O iAQQOiQ0b-b-t^b-COCOCOCpU5»O'^'^'«^CQCQC0Q0000"^ 



t* t* CO *(5 ^ ^4* ^4* ^ CO CO CO CO CO ^H ^ ^H O O O Oi Oft 00 00 ^ 
aOkOtOi/5ar^^^tO*OkOu5u5tOtcu3aoiOiOkO^^^^'<9* 



•^ 1 

«» i I 






s 



s 



»0»OiO*0«SM5>0"D*0«D^^^^^^^^^^'<r^ CO 



kO *0 lO ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ TT ^ ^ Cv CO 



»•• . ^ ^ ^ ^ ^ ^ ^ ^ ^ CO CO CO CO CO CO 



o 

09 



8 1 ^S^"5^5252S2^ 
I ^cocococOcOcOcoco 



Is 

I 



CO S S 



Oft . 

*0 i 



s 






§ 



^ 



9 



C? I 



s i 






THS JOUBITAI. OF BIOLGOICAL CBBiaSTBT, VOL. ZX, KG. 8 



366 



Nutritive Value of Proteins in Growth 






1' 



ssg 



s 






S 






-« ' 






00 



o o> 



:aj ?5 •-< 

00 00 X 
00 00 00 t>- 

— ' ^ ^ c^ 

Oft 00 00 t>> 
X X 00 

»-l 1-H O CO 
00 00 b- b- 

00 r^ r^ b- t^ 

"»Q~cc "o" eo~cr 
oo oo t^ r^ r* 

§8SJ2"J2~8S_S 

00 00 t^ t^ b- b- 50 



oc 

00 

a. 



00 

"ob" 

OS" 



o 






s 
■§ 



e«i 



T\ 



OiOOOOt^b-b-t^r^ 



So tO~^ "^<~C0 C^"rH~i5"' 



•^ I OO b* t>- r^ !» l>- O CO <0 CO 



o 



^" 



s 



■^ ooor^r^cocoupoGOt^ 



t»cococococococo^toto 



'5 



•"I COCOCOCO^O^^QiOtOkOUd 



H 

o 

n 






CO CO 

CO Ud 






s*^ 



to 



•s I i"^- 



•3; 






1 e^i ] 

1 »^ I 



S I »o 



S' s s s 
s s 



•^s 



IS 



I 



si 



6l C^ 









10 ^ ^ to ^ ^ 

^ SS uS ^ "(f* ^ 

& S iS 9~§l "5h 

"^Jl "^1 -^l ^^ ^^ ^^ 

^^ 'V ^J* ^ ^t cq 

^ ^ ^ ^ CO CO 

^ CO W W 00 



00 



I 00 



l>- 



g,'{:ggg'S"feSSSSS i S 



T3 



CO 



CO 
»o 



lO 

9 



Ilk 



•e* 

o 
.0 



o 



o 



s 



tS QQ 05 Ji J2 JS gS 

'IT CO CO CO CO CO CQ 



JJ? ;5I £? fe 

CO CO CO 01 



CO 

CO 
CO 



I > 



T. B. Osborne and L. B. Mendel 367 

tabic on page 366 on which a noniiai groN^-th was being made. 
These are not isolated instances. They have been repeated again 
and again and answer the obvious criticism that the failure to 
grow properly was always due to a deficiency in ingested calories. 
With our mode of investigation and its not inconsiderable 
limitations thus outlined, it is of interest, as has already been 
l^ointed out, to compare the relative efl&ciency of the same con- 
centrations of different proteins. Charts VI, VII, and VIII 
present results of this sort. The unlike efl&ciency of 12, 9, 6, 
4.5, and 2 per cent of the proteins casein, edestin, and lactalbu- 
min, is at once apparent. A content of protein which is utterly 
inadequate in the case of casein permits good growth when lact- 
ulbumhi is used. This, again, is not due to marked differences 
in the actual absolute consumption of protein, as the following 
(lata demonstrate : 



KA,T 





INITIAL 






INTAKE OF 


DIET 


BODY 
WEIQHT 

i 


GAIN 


Total 
j food 


Protei 




gm. 


gm. 


d»: 


gm. 


gm. 


f lactalbumin 


' 65 


99 


80 


; 589 


43.7 


ti 


43 


89 


80 


i 395 


29.3 


i • 


59 


91 


80 


575 


42.6 


>< 


60 


93 


80 


560 


41.5 


9'4 casein 


42 


59 


80 


435 


35.2 


9% cdcstin 


' 52 


60 


80 


465 


38.9 


Ai^'i lactalbumin 


i 42 


44 


80 


' 371 

1 


13.8 


i< 


41 


54 


80 


454 


16.8 


^' i casein 


i 59 


-3 


80 


1 366 


14.8 


4J' ( edestin 


83 


14 


80 


459 


19.2 


1 1 


65 


29 


80 


1 464 


19.4 



2115cf 

2123 cf 

2207 d' 

22109... 

20510^ 

21109 

•2044 d' 

20499 

2118^" 

21139 

21149 

It will be observed that the rats supplied with the 9 per cent 
casein or edesthi food ate quantities of protein approximately 
similar to those on the 9 per cent lactalbumin food, but gained 
only two-thirds ius much in weight during the same number of 
days. A comparison of the animals on the foods containing 4.5 
per cent of these proteins shows that while the protein intakes 
were of approximately the same order, the gains on lactalbumin 
were relatively much greater than hi the experiments with the 
other 4.5 per cent foods. 



368 



Nutritive Value of Proteins in Growth 



A more critical comparison of some of the proteins on the basis 
of their efficiency in permitting maintenance or promoting growth 
will be possible when further data are available. The facts 
adduced above correspond with observations already published 
by us to demonstrate that the proteins have unlike physiological 
values in supplementing the deficiencies of zein as a dietary 
protein.2* The economy of the different proteins as nutrients 
in growth appears to be closely bound up with their amino-acid 
make-up. 

APPENDIX. 

The composition of the foods used was as follows r^^ 

CA8EIK 



Protein 18 15 

Cystine 

Protein-free milk.' 28; 28 

Sucrose 

Starch 29,' 32 

Butter-fat 18| &-18 

Lard 7,19-7 



I 



0.36 



28' 28 

7I 7 

25' 18 

18| 10 
lOi 



per cent 



12 11.64 9 



9 6 4.5 4.0 2 

i 0.54, 0.54^ 0.135 

28' 28 ' 28, 28 '28 28 28 

9' 9 \ 10, 10 10 10 12 

29 28.46' 31 30.46' 31.5 31.365 32 



18 6-18 I 18 18 



719-7 



7 7 



18 
8 



18 

S 



18 
8 



LACTALBUMIN 



Protein 18, 11 

Protein-free milk. ... 1 28 28 

Sucrose 5 

Starch ; 28, 30 

Butter-fat 18, 6-18 

Lard 820-8 

Lactose 



per cent 

9 4.5 

28 28 
9 10 

29 31 
18 18 

7 8 



5 



28 
12 
31 
18 

8 



li.o 



5 



OVAL- 
BUMIN 

percent 



2 II 

28 28 

12 12 

32 33 

18 18 



9 ! 

28 
8' 
29 ' 
6-18 



S 8 20-8 

4 



OVO\ ITEI.LIX 



per cent 



9 f 

28 I 

8 I 

29| 

6-18. 

20-8 




2,S 
11 
27 
IS 


10 



*' Osborne and MendeJ: this Journal, xvii, p. 325, 1914; xviii, p. 1, 1914. 
2* These figures refer to the quantities of air-<lry foodstuffs; in the text 
the data for protein intake are calculated to a water-free basis. 



T. B. Osborne and L. B. Mendel 



369 



MILK 
PROTEINS 

percent 



Protein ! 12 

I 

Milk powder 46 

Protein-free milk 1 8 

Starch 19 

Butter-fat 5. 

Lard i 21. 



9 6 Protein. 



5 
5 



35| 25; Corn gluten 

11 15 Lactalbumin 

27 33! Protein-free milk 

7 9! Starch 

20 18' Butter-fat 

I Lard 



COBM 


OLUTKN 4- 


LACTALBUMIN 




per cent 


9 


6 


13 


' 8.8 


3 


2 


28 


; 28 


28 


33.2 


18 


18 


10 


10 



EDE8TIN 



28 



Protein 18, 

Lysine dichloride ' 

Na^CO, , , 

Protein-free milk [ 28: 

Sucrose , 

Starch ' 22' 

Butter-fat 18^ 6-18| 18 18.00 

Lard 14: 22-10, 10| 10.00 



per cent 

15 i 12; 11.46-11.76' 9 9' 

i 0.36- 0.81 ' O.8I' 

' 0.17-0.39! 0.39 

28; 28.00 i 28 ! 28 

i 7. 7.00 i 7-9 j 9 

29 ; 25,1 21.71-2-1.34: 28-29 28.8 



18 
10-7 



6^18 
18^ 



28 
10 
31 
18 
8 



28 
12 
32 

18 
8 



OLUTKXIX 

jter cent 

Protein 9 2 

Protein-free milk. . . . i 28 28 

Sucrose 1 6 12 

Starch ' 29 32 

Butter-fat G 18 

Lard 22 8 



OLYCINI.V 

per cent 

9 4.5 2 

28 28 , 28 
6 10 12 

29 81.5 32 
6 ,18 18 

22 8 8 



BQUASU-aSED 
OLOBULIN 

per cent 



QLIADIN 



per cent 



9 
28 

6 
29 

6 
22 



ZEIN 



gtn. 



4.5 
28 
10 
31.5 

18 
8 



9 
28 

7 
28 
18 
10 



2 
28 
12 
32 
IS 

8 



Protein 

Tryptophane 

Lysine dichloride. 

NajCOf 

Protein-free milk. 

Starch 

Butter-fat 

Lard 

Water 



18 


17.46 


16.92 




0.54 

1 


0.54 
0.81 
0.39 


28 


28 


28 


27 


^ 


26.34 


IS 


IS 


18 


9 


! 9 


9 


15 


1 ^^ 


15 



370 Nutritive Value of Proteins in Growth 




T. B. Osborne and L. B. Mendel 



37 • 




362 Nutritive Value of Proteins in Growth 

It is interesting to contrast not only the comparative rate of 
growth with different concentrations of the same protein (see 
Chart I for casein, Chart II for edestin, Chart III for lactalbumin), 
but also the unlike ability to grow on diets containing th'e same 
percentage of different proteins (Charts YL and VII). Criticisms 
which may be applied to this method of comparison will be dis- 
cussed later. 

Closely related to these features of nutrition are the compara- 
tive results obtained with the same concentration of different 
proteins when the nitrogenous intake is kept below the level at 
which adequate maintenance is possible. Even here the gradual 
fall of body weight is unlike with the different proteins (see 
Chart VIII). For further comparison Rats 2428 and 2445 were 
fed on diets otherwise similar, but containing only the exception- 
ally small amount of protein (0.63 per cent) present in the "pro- 
tein-free milk." 

The problem of the protein minimum. In studying the much 
debated question of the protein minimum animals have usually 
l)een fed with varying quantities of the mixed proteins char- 
acteristic of familiar food products. The foregoing discussion 
of our results with diets containing a single protein naturally 
suggests that the inequalities of the albuminous compounds m 
respect to their amino-acid make-up will give them quite unlike 
values when the minimum quantities for maintenance or for growth 
arc approached, the energy intake remaining the same. To cite 
an extreme case, no amount of zein food, however large, will 
enable rats to maintain their nutritive equilibrium. A small 
addition of tryptophane will at once convert the inefficient food 
into a maintenance ration. A rat has been kept without change 
of weight for more than six months on a diet containing zein and 
tryptophane as the sources of nitrogen (see Chart IX).^** If 

>^ Osborne and Mendel: Ztschr.f. physiol. Chem.^ Ixxx, p. 307, 1912. 
** Osborne and Mendel: this Journal, xii, p. 81, 1912. 
" Osborne and Mendel: ihid.^ xii, j). 473, 1912. 
'• Osborne and Mendel: ihid.y xv. p. 311, 1913. 
>^ Osborne and Mendel: ibid., xvi, p. 423, 1913-14. 
'" Osborne and Mendel: ibid., xvii, p. 401, 1914. 
' ' Osborne and Mendel: ibid., xviii, p. 1, 1914. 

-•* For othor data of this sort see Osborno and Mendel: ibid.y xvii, p. 
325, 1914. 



T. B. Osborne and L. B. Mende] 363 

a siDgle amino-acid, tryptophane, can play a rdle thus important 
in maintenance, it is more than likely that as the lowest limits of 
requirement are approached the inequalities in the proteins will 
make themselves noticeable in the unlike quantities needed for 
the different physiological performances. For example, with 
the other essential amino-acids equally well provided, the required 
minimum of the protein lowest in its yield of tryptophane may be 
expected to be greater than that of a protein comparatively rich 
in tryptophane precursors. 

The inequalities of different sources of protein in meeting the 
nutritive needs have been recognized in recent years by various 
investigators. In considering the comparative nutritive values 
of different isolated proteins one is at once confronted with the 
difficulty of comparing the nitrogenous needs of different annuals 
of different sizes and sexes at different ages. It is generally 
l^elieved that in growth a liberal supply of protein is required for 
constructive purposes beside what the wcar-and-tear functions 
call for. For the present we need not consider the added uses of 
protein as a physiological fuel material. How is the protein 
requirement of animals of unlike size to be measured? What 
unit shall serve as a basis for comparison? Evidently an older 
adult with an abundance of reserve fat cannot be contrasted gram 
per gram of body weight with a poorly nourished adolescent animal. 

In the experiments in this direction which we have thus far 
conducted the animals have had access to unlimited quantities 
of the mixtures of isolated food substances under investigation. 
The amount of food eaten, t.e., the total energy intake, has now 
been ascertained in a large number of instances of normal growth 
on such rations. The word "normar* is here used as synonymous 
with the average rate or curv^e of growth exhibited by the same 
species and sex living on unlimited quantities of a suitable mixed 
diet. To enable the reader to appreciate the range of variations 
in this matter the grams of food eaten per week by rats which 
were making normal growth have been ascertained at different 
stages of growth on the comparable diets given in the tables on 
pages 365 and 366. In each group the foods of the individual ani- 
mals differ in no respect except with regard to the protein, which is 
casein in some trials and edestin in others. 

The averages are expressed in graphic form in Chart X of the 



364 Nutritive Value of Proteins in Growth 

appendix. The early consistently smaller rate of food intake of 
normally growing rats on diets containing but tor-fat, of which 
the remarkable potency in facilitating growth has been dis- 
cu.«scrl ekewhere," is perhaps to be expected; for if a food is de- 
ficient in some essential element, it seems reasonable to suppose 
that the animal will endeavor to remedv the deficiency bv in- 
creasing its food intake up to the point where excess of food above 
the calorific and other requirements of the animal will bring 
alxiut the well known disturbances of digestion caused by over- 
eating, and consequently put an end to further increase of food 
intake. 

The range of variation in the tabulated results ou these strictly 
comparable diets is not inconsiderable; yet when one considers 
diflferenrres in the muscular activitv of the different individuals, 
they may after all not exceed what this variable factor would 
account for. At pre.sent this must remain mere conjecture. The 
ideal method of ascertaining the protein minimum would con- 
sist in feeding exactly equivalent amounts of energj- in the form 
of foods with imlike proportions of the individual proteins, thereby 
learning from the failure of proper gains where the minimum 
for growth lies, or what the limit of intake for maintenance con- 
sists in, when that is the function under consideration. Until 
recently it has been impracticable, if not impossible, for us to 
follow this procedure, although we now have experiments relat- 
ing to this problem in progress. 

In our observations on the rate of growth as well as the effi- 
ciency of maintenance when rats were fed on food mixtures con- 
taining a different content of protein, ranging from 18 per cent 
(which sufficed for adequate growth) to 2 per cent (which has not 
sufficed for maintenance even) (Charts I, II, III, VI, VII, VIII). 
the rate of food intake shows that the failures cannot usually be 
explained by a lack of enetgy in the diet. Frequently rats which 
failed to grow normally were eating sufficient food to enab'e them 
to grow if the protein factor had been adequate.^- In many 
cases the food intake exceeded the average of those given in the 

'* Osborne and Mendel: ibid., xvi, p. 423, 1913; xvii, p. 401, 1914. 
E. V. McCollum and M. Davis: ibid., xv, p. 167, 1913. 

*- For a discussion of the relation of growth to food intake, compare 
Hopkins: Jour, Physiol. j xliv, p. 425, 1912. 



T. B. Osborne and L. B. Mendel 



365 



I 

••a 
O 

■§ 

o 
.0 



to to C^ Oi 

b. t» b. CO 



oOb-b-b-b-iN.2525 



ssessssss 



s 



s 



O000O000b*.t^b*t»b«b«S>S5cO 



.V. 



I' 

o 



OOOOOOo5b*b*b*b«b*b«b«b«b«S5cOOcOcO 



b-b-b-F«-b-b-b-b-b-lN.b-t^b-o2>25oo35«5 



3 

U 

o 

00 I 

■§1 

■< : 






s 



00 . 

V i 



o 



< 
O 

u 

o 

a 
o 

n 



l^'^'^»-«000»OOQOOQlN.5D'^"^eOCOCOO^*-i050»OOt^C^ 

t»t>>t>>t»t>>t«coooococ0cococo<o<ococo>otOko>oic 



CO 



kOOdOdaocptou5tQup'^coeococ^«-i«H<-4QOt^b«t^M3csi 
t>>cococo^<oco;odcocD<ocDocococoickoio^kOko 



;e ■ 



10 

CO 
1 b- 



^ ! 



• b- 



CO 
b- 



I 

I 

. b» 

■ S 



lOiOiOiO-^COCOeOC^^-ii-ii-iOOSOib-CpiOiOCOCilQOi^ 

b-b»b-b-.i^b»b-t^b-b-i^b-i^cooo«5^^«5o35»o»c 



CO'^'^CfCO'-i'-iOOOOQOOOCOCOiQiQiQCOC^QOiQOQOO iCO 

b«b«b«bSb«t^b«COCOOOOOOcOOO<OOSkOii5iO^ '^ 



s \ 



CO 

CO 



0»b-upup'^CO»-«»-i»-iQO»0>OSOSOSQOb-COCO»0'^COC005<P OS 

cococococococococo^tOkOiOkOkOi^^kOkO»oub^u5-^^ ko 



Si M5QQOi05i^i^^-^-«o«o«S^"P"PTt!'^Tfe9cocoQoQQO^ MS 
S cococo&otokOkOkCu?^u2i^*ou^LOubtoubu5koioud^^^ &o 



S 



! "! 



l^t^CO»Q'^'^'^'^COeOCOCOCOi-i^^QQQOOOQQO^ ic^ ! 

io^kOk/5u'0iOkCkC^koiOko^kOU3kOi/5kC»o'^'^'9^^ !»o 






<S 



s 



J2 i 



b-COCOCOC<IN'M^»-«QOCib«.b-t^t^t^»OwpiOCOQOi 

io»ow5u5w5»o»o»o»o25'^'^'^"^"^'^'^TpTpTp"«#^co 



^ 



to *0 *0 ^ ^ ^ ^ ^ ^' ^9* ^9* ^' '9' '9' ^' ^ ^ ^^ ^^ ^ CO CO 



l^iO^»-^^QQQQQ000i^COCO»O 
■^ ^ "^ ^ '^ ^ ^ ^ ^ CO CO CO CO CO CO 



i i? 



§ 



o 

00 '< 

cs 

•2 j 
ft? I 



i«0 ^cocococococococo 



o 



CO S ^ 



{3 



8 ■ 



o 
a 

> 

< 



THB JOirBNAL OF KIOLOOXCAL GBKUISTRT, VOL. XX, NO. 8 



366 Nutritive Value of Proteins in Growth 



• 




g 


5 ac t C5 — 
X ai ci CC X 




s 






9. 

Ft 
9% 


© »3 c£ i.-; -^^x 
2 ^ X >! X 1^ 

* S » at 3 I- 

3 s s s s s 




1'. 

X 

■ »» 
X 

"?l 
J. 


fa. 




I 


M ?l ^ — Ci « 
X X X X t^ t^ 




s 






1 


55 !5 ^ Ci :S "O ri 
cS dC X i« i« 1^ f- 




'2 


0^ 




I 


8SS5Si2Ss 




X 


•0 

5 




9 

CI 


issssi^i^g^s 




X 


^ 






i 5 S a rt ;t f^ :: s 

cb X X t« r» K- i« t« 2: 


— 




••- 


X 






g" 


55 lO -f ^ eo d — iQ" 

QC X l» l» l« l>- t« t» 




1^ 


c 

6. 




s 


CJ 10 CO C»l ^ ^ Ci l>- t>- ^ 
X !» I» i» l» !» ;S ^ :S 




X 

• 


5 

"I. 


3 

m 



% 


>. CO x 1^ I-. ;5 "^ 

l>- i» i» CO cc c5 c^ 




"5 




^" 


CO X to t« CO tft X 1^ 

1^ t« «$ CO CO CO u? t^ 




i^ 



X 

s 


'A 

< 

1 




• 

t>- l>- CO CO ri CI C»l ^ Q -^ "^ 

1^ CO ^ CO CO CO CO CO ub u*? 
SococoSSSi^OkotoS 


lis 

CO 

• 

Ci 


%* 


s 


5 


^« ri Q ci~"cio bi t<- CO "-^ CI 

CO t^ >0 >i) kC u3 iC 


t-H 






■5 



n 






55 
5 


1—" 


1 




. 


ii5"5»o25'^W'^^'^^^ 


I—" 


SS 


: 
E ' 




8| 




5' 

■ CO' 


^ 1 




S 


^J* ^T ^1* ^t* "T ^ ^ CO CO CO CO 




' s 






l-» 


^ n CO CO CO ro co 




CO 
CO 


niake 




8 

■ 


ro w CO S 




CI 

CO 


^ood 11 
















c! 

u 

> 




T. B. Osborne and L. B. Mendel 



367 



table on page 366 on which a normal growi;h was being made. 
These are not isolated instances. They have been repeated again 
and again and answer the obvious criticism that the failure to 
grow properlj^ was always due to a deficiency in ingested calories. 
With our mode of investigation and its not inconsiderable 
limitations thus outlined, it is of interest, as has already been 
pointed out, to compare the relative efficiencj' of the same con- 
centrations of different proteins. Charts VI, VII, and VIII 
present results of this sort. The unlike efficiency of 12, 9, 6, 
4.5, and 2 per cent of the proteins casein, edestin, and lactalbu- 
uiin, is at once apparent. A content of prot^jin which is utterly 
inadequate in the case of casein permits good growth when lact- 
ulbumin is used. This, again, is not due to marked differences 
in the actual absolute consumption of protein, as the following 
data demonstrate: 



I 



KAT 



2115cf 
21230^ 
2207 d' 
22109 
2051d^ 
21109 

2044 c/' 
20499 
2118c:f 
21139 
21149 



DIET 



9 / lactalbumin 



9' I casein 
9% edestin 

^^ /c lactalbumin 



u 



4i' I casein 
4i' f edestin 



I 

INITIAL' 
BODY I 

weight! 

I 



gm. 

65 
43 
59 
GO 
42 
52 



U.UN 



gm. 

99 
89 
91 
93 
59 
60 



42 44 

41 54 

59 i -3 

83 ; 14 

65 29 



80 
80 
80 
80 
80 
80 

80 
80 
80 
80 
80 



intake or 

7^1 Protein 
gm. gm. 

589 : 43.7 
395 I 29.3 
575 i 42.6 
560 41.5 
435 35.2 
465 38.9 



371 
454 
366 
459 



13.8 
16.8 
14.8 
19.2 
19.4 



It will be observed that the rats supplied with the 9 per cent 
casein or edestin food ate quantities of protein approximately 
similar to those on the 9 per cent lactalbumin food, but gained 
only two-thirds tis much in weight during the same number of 
days. A comparison of the animals on the foods containhig 4.5 
per cent of these proteins shows that while the protein intakes 
were of approximately the same order, the gains on lactalbumin 
were relatively much greater than hi the experiments with tlie 
other 4.5 per cent foods. 



368 



Nutritive Value of Proteins in Growth 



V 



A more critical comparison of some of the proteins on the basis 
of their efficiency in permitting maintenance or promoting growth 
will be possible when further data are available. The facts 
adduced above correspond with observations already published 
by us to demonstrate that the proteins have unlike physiological 
values in supplementing the deficiencies of zein as a dietary 
protem.^ The economy of the different proteins as nutrients 
in growth appears to be closely bound up with their amino-acid 
make-up. 

APPENDIX. 

The composition of the foods used was as follows i^^ 

CASEIN 



Protein 18 15 

Cystine 

Protein-free milk. 28, 28 

Sucrose 

Starch 29 32 

Butter-fat 18^ 6-18 

Lard 719-7 



12' 11.64 
I 0.36 
28| 28 
7i 7 
25i 18 
18; 10 

lo! 



per cent 

9 9 6 4.5 

' 0.54 0.54 

28-28 '28, 28 28 

9, 9 10 10 10 

29, 28.46 31 30.46 31.5 

18, 6-18 ' 18 18 , 18 

719-7 7 7 8 



4.5 2 

0.135 

28 2S 

10 , 12 

31.3<>;'5 32 

IS IS 

8 8 



I 
Protein ; 18 11 

Protein-free milk. ... 28 28 

Sucrose ' 5 

Starch 28^ 30 

Butter-fat 18| 6-18 

Lard 820-8 

Lactose 



LACTALBUMIN 



per cent 



9 4.5 
28 28 

9 10 
29. 31.5 
18 18 

7 8 



2 

28 
12 
31 
18 

8 



If.o 



OVAL- 
BU.MIX 

percent 



2 II 
28, 28 
12 12 
32 33 
18 18 



9 

28 

8 

29 

6-18 



8 8 20-8 



OVOVITKI.MV 



per c*tit 



9 ! 

28 • 

8! 

29;' 

6-lS 

20-8 




2N 
11 
27 
l.S 


10 



*' Osborne and Mendel: this Journal, xvii, p. 325, 1914; xviii, p. 1, 1914. 
^* These figures refer to the quantities of air-dry foodstuffs; in the text 
the data for protein intake are calculated to a water-free hasis. 



T. B. Osborne and L. B. Mendel 



369 



MILK 
PROTBIK8 



percent 



Protein i 12 

I 
I 

Milk powder j 46 

Protein-free milk i 8 

Starch ! 19 

Butter-fat l 5.5 

Lard 21.5 



9 6 Protein. 



35 25i Com gluten 

11 15^ Lactalbumin 

27 33 Protein-free milk 

7, 9 Starch 

20 18 Butter-fat 

Lard 



CORN OLI7TKM + 
LACTALBUMIN 



per cent 



9 



13 


8.8 


3 


2 


28 


28 


28 


33.2 


18 


18 


10 


10 







EDBSTIX 



per cent 



Protein 18 15 

Lysine dichloride 

XazCOa 

Protein-free milk 28! 28 

Sucrose 

Starch 22 29 

Butter-fat , 18; 6-18 

Lard 14 22-lOi 



12 



11.46-11.76 9 9 4 

, 0.36- 0.81 0.81 

' 0.17-0.39i 0.39 

28 28.00 j 28 ' 28 28 

7 7.00 7-9 9 

25 21.71-24.34 28-29 28.8 

18 18.00 18 6-18 

10 10.00 : 10-7 18-6 



10 
31.5 

18 
8 



28 
12 
32 
18 
8 



OI.rXKN'IN OLYCINiy 

;>«T cent per cent 



Protein 2 

Protein-free milk. ... 28 28 

Sucrose 6 12 

Starch 2^) 32 

Butter-fat G 18 

Lard 22 8 



SQUASH-REED 
GLOBULIN 

per cent 



9 4.5 2 

28 28 28 i 
6 10 12 

29 81.5 32 
6 18 18 



22 8 



8 



9 
28 

6 

29 

6 
00 



4.5 
28 
10 
31.5 

18 
8 



OLIADIN 



per cent 



9 

28 
7 
28 
18 
10 



2 
28 
12 
32 

is 

8 



ZEIN 



Protein 

Tryptophane 

Lysine dichloride 

XajCO. 

Protein-free milk. 

Starch 

Butter-fat 

Lard 

Water 





gin. 




18 


17.40 


16.92 




O.M 


0.54 
0.81 
0.39 


28 


28 


28 


27 


27 


26.34 


IN 


IS 


18 


9 


9 


9 


15 


15 


15 



370 Nutritive Value of Proteins in Growth 

























/ 
















































^ 




J 


/ 




















:g 




h 






















■« 


; 


■i 






















J 


^ 


/ 


"* 
























\ 


1 






^^ 














■^\ 




\ 


/ 








-Xi 














\ 






\ 








.\ 


s 




\ 




s 




^1 




N 


\ 






"" 


\ 


-s 


\ 




\ 


^ 


\ 




V 


^ 




'> 




\ 


\ 


1 


\ 








N 






^ 


-V 


S 






^ 




1 








^ 










■% 


s, 


K 


^ 




'^ 










^ 


\, 






^ 


^"^j 


\ 


>< 


5^ 
















^ 


^ 


>= 


\ 


-■ 






















s^ 


x 




























>fe 


\ 





























|5- 

£•5 



T. B. Osborne and L. B. Mendel 

























, 




















/ 






/ 


















« 


/ 






1 




















\ 






// 






















\ 




// 






















\ 




iJ 






\ 












^ 






^ 


J 






















\ 




/ 






\ 






\ 






5 
















^^ 






\ 




1 








r" 










\ 




\ 




g 








\ 












\ 


1 


\ 


\ 




% 






\ 


■^ 






,^ 


\ 


\ 




\ 


\ 


/ 










^ 


\ 


\ 




\ 


\ 


s 


^ 
















\ 


s - 




\ 


\ 


\ 


\ 














\ 


V 

\ 


^. 




V 







\ 
















\ 


■5? 


i^- 


■^ 




^ 




















V 


/ 


^ 




^ 












^ 








> 


^ 


'aiSJg 



.a .a 
II 



II 



Nutritive Value ot Proteins in Growth 




374 Nutritive Value of Proteins in Growth 




376 Nutritive Value of Proteins in Growth 




-lociflvj t-' 



(/iiAKT VII. Comparison of growth on diets containing approximately the same 
percentage (4.5 per rent) of diflfcrent proteins; namely, lactalbumin, edestin, caseirif 
globulin (squash-seed), and glycinhi (soy bean). The composition of the foods used 
is given in the tal)leH in the appendix. 



T. B. Osborne and L. B. Mendel 




Chart VIII. Comparison ot changes in body weight induced by Approximately 
the same inadequate percentage (2 per cent) of lacfalbumin, edeslin, casein, glutenin 
(vheat), glycinin (soy bean), and gliadin (nhcat) in the diet. For contrast the 
effect of a practically protein-free diet is shown id Rats 1950, 2443, and 2428. The 
composition of the foods used is given in the tables in the appendix. 



378 



Nutritive Value of Proteins in Growth 






iL. 



i o 



fi 



1 




3c(UL^;i 



oT i: .22 £ -5 




tophan 
one-ha 
pite th 
and tl 
ods use 










** h C .c 










o O O cj fl 










a -i ^ 2 1 




£ e §■*! 




^ '^ ^ CO O 




•3 3 g:5 ^ 




s f^-fev. § 




S N . O X 




2 S +-« ^ 




5* s 2^ 




.5 ^ 'S « H 








^ -s « 








o '3 o 2 'S 










and 
eigh 
not 
a mi 




§ s r ss 




decline 
ptophai 
in bod> 
grow w 
lit si/.e 


• 


iS •** t* "^ ** 







08 


9 O '^ ''***^ 




«^-c « § p 




Showinj 
^idition 
significac 
et the ca 
tely grew 
tables in 


. • 05 •-* oa 


O 




X 

i-*^ 


^^ O O Cj •-* 




Chart 
owth aft 
ar withi 
adcquat 
limal ult 


a 
a 




^^B § 


00 



3 -•* 




FURTHER OBSERVATIONS OF THE INFLUENCE OF 

NATURAL FATS UPON GROWTH.^ 

By THOMAS B. OSBORNE and LAFAYETTE B. MENDEL. 
With the Cooperation of Edna L. Ferry and Alfred J. Wakeman. 

(From the Laboratory of the Connecticut Agricultural Experiment Station 

and the Sheffield Laboratory of Physiological Chemistry in 

Yale University J New Haven.) 

(Received for publication, January 30, 1915.) 

When young albino rats are fed on a ration composed of isolated 
and purified protein, a carbohydrate like starch, inorganic salts 
furnished in the form of what we have termed ** protein-free milk"* 
(which is likewise fat-free), together with commercial lard, they 
usually grow normally for about three months, but never attain 
their full size. Sooner or later nutritive disaster manifests itself 
by a complete or partial cessation of growth which ultimately 
(and sometimes precipitately) ends in a decline in body weight, 
followed by death if a suitable change in diet is not promptly 
instituted. The disturbance of growth may be attended by 
symptoms of malnutrition (such as infections of the eye) giving 
evidence of impaired resistance to bacterial invasion. Recovery 
promptly follows the substitution of a part of the lard of the 
ration by certain other natural fats.' 

* The expenses of this investigation were shared by the Connecticut 
Agricultural Experiment Station and the Carnegie Institution of Washing- 
ton, D. C. 

* T. B. Osborne and L. B. Mendel: Feeding Experiments with Isolated 
Food-Substances, Carnegie Institution of Washington, Publication No. 156, 
pt. ii, p. SO, 1911; see also Ztschr. f. physiol. Chew., Ixxx, pp. 315-16, 1912. 

' E. V. McCollum andM. Davis: this Journaly xv, p. 167, 1913. Osborne 
and Mendel: ibid., xv, p. 311; xvi, p. 423, 1913-14. McCollum and Davis: 
Proc. Soc. Exper. Biol, and Med., xi, p. 101, 1914. Osborne and Mendel: 
this Journal, xvii, p. 401, 1914. McCollum and Davis: ibid., xix, p. 24o, 
1914. 

The word "fat'* is used in this paper to refer to the mixtures which 
ordinarily are included by that designation, rather than to the pure 
glycerides of fatty acids. 

379 



38o Influence of Natural Fats upon Growth 

The importance of this feature of the dietary in the use of 
mixtures of isolated food substances as the exclusive ration of 
growing animals, is emphasized by experience which we have 
accumulated during several years. Rats have been raised and 
bred, into the third generation, on diets consisting of: 

per cent 

Edestin 18 

Starch 24 

''Protein-free milk'' 28 

Commercial lard 12 

Butter-fat 18 

Some of our animals are still living after 620 days on such a 
ration. In other words, when the fat component of the diet is 
suitable from the outset, the peculiar nutritive failures described 
do not put in their appearance. This lends confirmation to the 
conclusion that the "standard" diets containing lard or certain 
other substances as the sole fat admixture are unquestionably 
deficient in some respect. 

The diflference in the eflfects on growth which characterize the 
different natural fats investigated suggested that perhaps the 
inefficiency of the lard was due to the technical processes to which 
the fat of the pigs was subjected in preparation for the market. 
Lard was therefore prepared in the laboratory by the following 
procedure from fat brought directly from the slaughter house: 

The crude material was comminuted, and filtered through paper at a 
temperature just above its melting point. Overheating was prevented 
by putting the lard in a large filter placed in an oven at a temperature just 
sufficient to melt the fat and allow it to pass slowly through the paper. 

With this "laboratory lard,*' used in place of the commercial 
product in the diet, without other fats, the cessation of growth 
occurred quite as promptly and characteristically as before 
(see Chart I); and restoration was accomplished by the usual 
method of adding another natural fat (such as butter-fat) to the 
diet to replace part of the lard. Other facts are in further har- 
mony with the conclusion that the inefficiency of the lard is not due 
to the heating to which the fat may have been subjected in its 
preparation. Almond oil, likewise, which failed to restore growth 



T. B. Osborne and L. B. Mendel 381 

after the declines on lard diets,* was prepared in the laboratory 
without the aid of heat, by grinding the nuts, pressing out, and 
filtering the oil at room temperature. On the other hand, butter- v{^ 
fat through which live steam was passed for two and one-half 
hours or longer did not lose its characteristic restorative prop- 
erties (see Chart I, Rats 1962 and 1976). McCoUum and 
Davis have recently obtained results indicating that the sub- 
stance, or substances, present in butter-fat which exert such a 
marked stimulating action on growi-h, are sufficiently stable 
to withstand conditions of saponification which they have em- 
ployed.^ Accordingly, certain fats may retain their efficiency in 
this regard despite vigorous treatment with heat^ or chemical 
agents; whereas other fats remain practically inefficient even after 
the most gentle manipulation. ^^ 

To the Hst of natural fats found to contain the factor efficient 
in facilitating growth on the diet outlined above we can now add j 
beef-fat. " The material was prepared as follows: 

Abdominal fat from cattle was cut up, and filtered through paper at a 
temperature just above its melting point. Higher temperatures were 
avoided by conducting the filtration as with lard (see page 380). 

In Chart II are reproduced some of the growth curves of rats 
fed on a ration of: 

per cent 

Casein 18 

Edestin 18 

Starch 29 20 

*' Protein-free milk" 28 28 

Commercial lard 7 16 

Beef-fat 18 IS 

In some cases the decline invariably experienced in about three 
months when lard alone was used has been averted by admixture 
of beef-fat. The recoveries when beef-fat is fed after decline on 
the lard diets (Rats 1914, 2130, 1924, Chart II), and the continu- 
ance of growth when beef-fat feeding is begun very early, are not 
as marked or prolonged as are those accomplished by l)utter-fat 

* Osborne and Mendel: this Journaly xvii, p. 401, 1914. 

* McCollum and Davis: ihxd.^ xix, p. 245, 1914. 

* In accord with this, McColium and Davis {Proc. Soc. Exper. Biol, and 
Med.f xi, p. 101, 1914) have also found extracts of boiled egg to be effective. 

THK JOURNAL OF BIOLOGICAL CUEMIBTKT, VOL. XX, NO. 3 



382 Influence of Natural Fats upon Growth 

additions. Failure has in most cases ultimately ensued (see 
Rats 1793, 1914, 1924, Chart II, and 2349, Chart III) and res- 
toration been effected by the use of butter-fat or "butter oil." 
These failures have, however, only been not^d after a far longer 
period than occurs where lard alone is used. Chart III shows 
the growth of rats supplied with beef-fat in the diet from a very 
early period. On this ration growth is continued for a far longer 
period than on the diets containing lard only. Growth has not, 
however, continued to complete maturity as it does with butter- 
fat in the food. For the present we interpret this as indicating 
that the content of the growth-promoting substance is so small 
^that it ultimately becomes quantitatively inadequate. 

Reports hitherto pubUshed include the following fats that have 
been tested from the standpoint considered in this paper :^ with 
positive results — ^butter-fat,^** egg yolk fat,**' cod liver oil;^° with 
negative results — ^lard,*-* oUve oil,® almond oil.^^ 

Concentration of the growth-promoting substance. The con- 
centration of the effective substance contained in butter-fat and 
in beef-fat has been attempted by the following process: 

A large quantity of absolute alcohol is saturated at 40^ with the melted 
fat and then cooled to —15^. After standing in the freezing mixture for 
some time to allow the higher melting fats to crystallize, the solution is 
rapidly filtered on a large Buchner funnel. The clear alcoholic solution 
is then concentrated at 40® in vacuo until the alcohol is all removed. The 
clear, deep yellow oil which remains constitutes the ** butter oil" (Fraction 
III) or *'beef oil" used in these experiments. 

. The substance designated *' Fraction I" was prepared by recrystalliz- 
ing the part of the butter-fat filtered out as above described, until the 
glycerides liquid at room temperature were removed and a snow-white 
crystalline solid product obtained, which was equal to a little more than 
one-half of the butter-fat taken. 

As a test procedure for the identification of "active'' material 
we have employed the plan either of adding the substance to be 

'An investigation of the behavior of the substance in butter-fat which 
exerts a growth-promoting influence is being conducted with reference 
to its diflfusibility through rubber membranes, by Professor Gics at Colum- 
bia University. 

' Osborne and Mendel: this Journal, xvi, p. 423, 1913. 

•McCollum and Davis: ibid., xv, p. 167, 1913. 
** Osborne and Mendel: ibid.y xvii, p. 401, 1914. 



T. B. Osborne and L. B. Mendel 383 

tested to the diet of rats which have failed on the lard ration 
(see above) and watching for a resumption of growth, or of pre- 
venting the anticipated failure by an early addition of the in- 
vestigated fraction to the ration. 

Our experiments indicate clearly that the effective substance 
tends to concentrate in the fractions which we may term " butter 
oil" and "beef oil," respectively; i.e., in those fractions which do 
not crystallize during the process described under the mode of 
preparation. The harder butter-fats (Fraction I) have failed 
to give evidence of growth-promoting properties (see Chart IV), 
in contrast with the liquid oil fraction (Fraction III) which 
promotes recoveries after failure on Fraction I (see Chart IV). 
That the groAvth-promoting substance of the beef-fat has been 
concentrated in the beef oil is shown in Chart V. Here again 
the purulent condition of the eyes, hitherto noted with the lard 
diets, also occurred with diets containing the inefficient fractions 
of the butter-fat and were cured by supplying the butter oil, 
as happened when the entire butter-fat was given. Whether 
these fat fractions deteriorate with age and exposure remains to 
be learned. 

Preliminary to the determination of the quantities of "active" 
substance involved in the growth-promoting phenomena which 
have been recorded for the naturally occurring fats, we have made 
some investigation of the proportions of butter-fat, butter oil, 
cod hver oil, and beef oil needed to avert decline. Among hun- 
dreds of rats tested on the inadequate lard diets, one animal con- 
tinued to thrive for the exceptional period of 252 days before the 
inevitable failure ensued; here too, however, addition of butter-fat 
induced a return of nutritive equilibrium. It is important to 
mention this unexpected endurance, unusual though it is, in 
evaluating the results observed with small additions of the 
effective fats. Our "standard" diets have as a rule contained 
18 per cent of the latter. Many rats have been grown success- 
fully on a diet containing 6 per cent of butter-fat; some on 3 
per cent; a few on 1 per cent only. That these lower quantities 
approach the limits of adequacy is evidenced by the ultimate 
nutritive failure and the sul)sequent improvement following mere 
increase in the fat investigated. A few typical cases of such 
"quantitative" failures are recorded in Chart VI. 



yf 



^ 



384 Influence of Natural Fats upon Growth 

The proportions of added "butter oil" and "beef oil" fractions 
selected were usually 6 per cent of the entire food, obviously 
representing a much larger addendum of the original fat. We 
have already shown that 6 per cent of cod liver oil also is satis- 
factory for growth." 

In comparing the numerous records of growth on diets contain- 
ing butter-fat and beef-fat, respectively, we have gained the 
impression already referred to, that butter-fat is more effective 
in permitting growth than equivalent quantities of the beef-fat. 
Recoveries are less prompt and prolonged growth is less satis- 
factory when the latter is used. In this connection it may be 
observed that the yield of the liquid "oil" fraction from butter- 
fat is considerably larger than that from beef-fat. The findings in 
respect to the beef-fat explain the fact, which we have observed, 
that commercial oleomargarine also effects recovery in rats that 
have declined on the lard diets. 

The new features of this communication may be summarized 
as follows: 

The failure of lard to promote growth in the same manner as 
other natural fats (i.e., butter-fat, egg yolk fat, cod liver oil) do, 
is not attributable to deteriorating changes arising from heat or 
chemical agents in the commercial manufacture of the product. ! 

Heating butter-fat with steam does not destroy its growth- 
promoting efficiency. 
f "Beef-fat "idfiQ renders the inefficient diets used by us more suit- 
able for producing growth in rats than does lard. 

When butter-fat and beef-fat are subjected to fractional ( 

crystallization from alcohol, the growth-promoting factor remains 
in the mother liquor or "oil" fractions. The fractions containing 
the fats with high melting points are ineflfective. 

Some quantitative aspects of the growth-promoting efficiency j 

of the natural fats are discussed. j 

"Osborne and Mendel: ibid.y xvii, p. 401, 19H, Chart III, Rats 1893 
and 1898. 






T. B. Osborne and L. B. Mendel 



385 




Chart I. Ceaaation of growth and nutritive failure on diets containing "laboratory 
lard" Bs the sole (at. Reatoratioa of growth by replacing part of the lard witb 
egK-fal, butt«r-rtit, "butter oil," or commercial oleomargarine. 

The food mixtures consisted of: 

vrrent 

Ptslain 18 

Pratein-lrae milk £j 

Surah I*-W 



Vi* 



<r 



>^> 



r^ 






-'^ 



4 









/ 



/ 



— ^jJ^Su^ ~ 

~^ ~ » -~~ — -^ T^~" 

• 



; '^ i 



'-i 



b_ 









JB 

X 



■- a 







JO 



f: 




t 

S 



* ^ z: 



9 



T. B. Osborne and L. B. Mendel 



387 



240r 



no 



200 




Chart III. Growth on diets containing- beef-fat in the following 
mixture: 



Casein 

Protein-free milk. 

Starch 

Beef-fat 

Urd 



f.fr ctnt 
IS 
28 
29 
18 



The eflfect on growth is far better than in the case of the lanl diets; 
but it is inferior to that obtained with the butter-fat rations. 



T. B. Osborne and L. B. Mendel 



387 



240r 



KO 



200 




Days 

Chart III. 
mixture : 



Growth on diets containing- bccf-fat in the following 



Casein 

Protein-free milk. 

Starch 

Beef-fat 

Lard 



j.(T cent 
IS 
28 
29 
IS 
7 



The eflfert on growth is far better than in the case of the lard diets; 
but it is inferior to that obtained with the butter-fat rations. 



588 Influence of Natural Fats upon Growth 




Chart IV. Showing cejaation of growth oa diets coutuining lard und the harder butt«r 
[ftt Fraction I, Tolloved by recovery and renewal of growth when the "hutlcr oil" Fractia 
III was fed. 

The diets consisted of: 



Pniteln-IcM milli 

l^rd (vithor vithont buttfr-fmt tnctioni). 



T. B. Osborne and L. B. Mendel 



389 



iSO\ 




40*- 



Oays 



Chart V. Showing the growth-promoting properties of "beef oil" 
added to the food after cessation of growth on lard diets. The food 
mixtures resembled those described on other charts. 



THE NON-PROTEIN NITROGENOUS COMPOUNDS OF 

THE BLOOD IN NEPHRITIS, WITH SPECIAL REF- 

ERENCE TO CREATININE AND URIC ACID. 

By victor C. MYERS and MORRIS S. FINE. 

{From the Laboratory of Pathological Chemistry ^ New York Post-Graduate 

Medical School and Hospital, New York.) 

(Received for publication, January 31, 1915.) 

Since the introduction of simple methods for the determination 
of the total non-protein and urea nitrogen of the blood by Folin 
and Denis, a considerable literature has accumulated on the sub- 
ject. Nephritis has been the pathological condition to receive 
special investigation, since it is in certain types of this disease 
that a retention occurs. Aside from the observations reported 
by Folin and Denis, however, practically no data have l^een 
presented in this connection dealing with the accumulation of 
creatinine and uric acid in the blood. Their data represent iso- 
lated observations on a number of cases. In the present paper 
consideration has been given especially to this phase of the ques- 
tion, but the study was made as intensive as conditions would 
permit on a few selected cases. 

The non-protein and urea nitrogen of the blood has been 
found to fall within normal limits in many cases of nephritis. 
In those cases tending toward uremia, however, the values are 
increased and may reach figures of 350 mgm. per 100 cc. for 
the non-protein nitrogen, and 300 mgm. for the urea nitrogen.^ 
The series of observations recently reported by Tileston and Com- 
fort for both the non-protein and urea nitrogen and those by 

» O. Folin and W. Denis: this Journal, xiv, p. 29, 1913; xvii, p. 487, 1914. 
C. B. Farr and J. H. Austin: Jour. Exper. Med., xviii, p. 228, 1913. J. H. 
Agnew: Arch. Int. Med., xiii, p. 485, 1914. F. C. McLean and L. Selling: 
this Journal, xix, p. 31, 1914. C. Frothingham, Jr., and W. G. Smillie: 
Arch. Int. Med., xiv, p. 541, 1914. W. Tileston and C. W. Comfort, Jr.: 
ibid., p. 620. C. B. Farr and E. B. Krumbhaar: Jour. Atn. Med. Assn., 
Ixiii, p. 2214, 1914. 

39^ 



392 Non- Protein Nitrogenous Compounds of Blood 

Frothingham and SmiUie for the non-protein nitrogen are especi- 
ally illuminating on this subject. 

As might be expected, cases of uremia are accompanied by an 
accumulation not only of urea, but also of uric acid and creatinine, 
as is well illustrated by the recent studies of Folin and Denis.* 
We^ had already taken up this phase of the problem previous to 
the appearance of the paper by Folin and Denis; and at about the 
same time Neubauer,* in studying the impaired ability of the 
nephritic kidney to eliminate creatinine, reported a marked re- 
tention of creatinine in the blood in a case of uremia. A few 
observations on the creatinine content of dog's blood under dif- 
ferent experimental conditions have been reported by Shaffer.* 

Methods employed. 

The total solids were determined by collecting 0.3 to 0.6 of a 
gram of blood on a block of pressed filter paper and suspended 
by a wire hook from the stopper of a specially constructed weigh- 
ing bottle. The Kjeldahl method was employed for the total 
nitrogen, while for the non-protein nitrogen, uric acid, creatinine, 
and creatine the technique was, with slight modifications, that 
described by Folin and Denis. In the urea estimations the con- 
centrated urease® described by Van Slyke and CuUen was employed 
on either the fresh blood or residue of the same alcoholic filtrate 
• employed for the non-protein nitrogen. The ammonia thus 
obtained was aerated and subsequently determined colorimetri- 
cally as in the non-protein nitrogen estimation. The chlorides 
were determined by titration after removal of the protein (by 
coagulation in the presence of ,^(y acetic acid and subsequent 
treatment of the filtrate with a few drops of colloidal iron to 
remove the last trace of protein). A detailed description of 

' Folin and Denis: this Journal^ xvii, p. 487, 1914. 

* V. C. Myers and M. 8. Fine: Froc. Soc. Exper. Biol, and Med., xi, p. 
132, 1914. 

* O. Neubauer: MUnchen. m^d. Wchnschr., Ixi, p. 857, 1914. 

* P. A. Shaffer: this Journaly xviii, p. 535, 1914. 

« D. D. Van Slyke and G. E. Cullcn: ibid., xix, p. 211, 1914. We are 
indebted to Dr. I. F. Harris of the Arlington Chemical Company for our 
supply of urease. 



V. C. Myers and M. S. Fine 393 

the technique we have employed in blood analysis may be found 
elsewhere.^ 

Twenty-five cases^ are included in the present series, the first 
nine of which are of the retention type of nephritis. In two of 
these nine cases edema was present. The first three of these 
cases tabulated are of special interest since they are pronounced 
illustrations of different types of retention. The few cases other 
than nephritis are tabulated for comparison. 

Case 1, suffering from mercuric chloride poisoning, showed a 
very pronounced retention of all the non-protein nitrogenous 
constituents, the figures for non-protein nitrogen, urea, uric 
acid, and creatinine being decidedly higher than in any case 
reported by previous workers. No urine was passed for the first 
five days and no appreciable amount for the first ten days. After 
decapsulation of the kidneys on the sixth day the renal activity 
appeared to improve, and at one time it had sufficiently recovered 
to cause a reduction in the concentration of the creatinine from 
33.3 to 14.8 mgm. per 100 cc; but with the decline of the patient, 
the kidneys became less active and the creatinine again increased. 
The quite favorable output of total nitrogen was insufficient at 
any time, however, to reduce materially the non-protein and urea 
nitrogen of the l^lood, despite the favorable influence on the uric 
acid and creatinine. Although the highest concentration of uric 
acid and creatinine, as well as very high figures for non-protein 
and urea nitrogen, were found on November 20, uremic symptoms 
were not observed until a week later. This would seem to lend 
support to the current view that uremia is not a result wholly 
of the retention of these nitrogenous waste products.® 

(^ase 2 is interesting as illustrative of the condition of the blood 
and urine in a very severe case of interstitial nephritis with uremia 
but no ed(*ma. In this case there was a gradual decrease in the 

" Papers in The Posl-GraduatCj 1914-15, and collected as The Chemical 
Composition of the Blood in Health and Disease. 

•» Practically all the cases here reported were patients in the medical 
wards of this Hospital. We are indebted to the Director of the Depart- 
ment, Dr. Quintard, to Drs. Chace, Kast, and Halsey for many courtesies 
extended in connection with these studies; to the resident physician, Dr. 
W. G. Lough, and to Dr. F. D. Gorham for aid in following the cases and 
arranging the case hi.stories. 

» See F. W. Poahody : Arch. Int. Med., xiv, p. 23G. 1914. 



00 

I 






s 



' •0«d •» 8»»wid«oq J 



i 



ra ^ S ci ^ 3 S 

d d d <-i C4 1-H o 



10»N w WPHOWO 



oai^vaiQ 



O 1^ C« 1-1 1-^ o o 



3 H 

So 

p 



:a* C^ OJ C^ "3 C3 00 

p ^ 1-; CO '^P "^1* c<j 
d d d d d d d 



auiai)«ojo 



pi J« OUQ 



MI«V)X 



X^iA«iS ogioodg 



oumiojv 



Ok 



s s; s S8.S s s 

G G <6 G ^ G G 

1-; Ci CO -^ CO « 

c> G <6 G <6 G <6 



C9 O) CO o CO r^ 00 

r<« CO 00 1-j 00 i>- wj 

d c^ -^ d r^ d Tj* 

■54 ^H i-H 1-H »-^ «-H «-H 

O O O O O O O 



b 



gr^ Q r^ 00 o« c^ 
00 O 9 00 t^ «^ 

1^ ^ 25 o 1-I 00 i>- 






o rJ. ct »i -A d 'T 

.-H rH 1-1 c^ c<i c5 ^ 

• •••••. 

o o o o o o o 

^ ;z; ;z; ;5 ;z; ;z; Q 



8« upuoiqo 



oui)«ajo 



oaiai^vojQ 



Ok 



r^ CO Q rH w 

kO ^ o to o 

d d d d d 



CO 



S 

rs 



5 

o 

s 



P«w 31 JQ 



N «"n 



lo^oid-iiot^ 



^ i 

8 ' 



M l»^oj. 



■pnos i»»ox 









•^ C« -^ W 1-1 d 

2S2a S 8 

r*. o CO 1-j 00 1-j 

t> c^ CO 00 -^ r^ 

Cl CO CO C^ «-< 1-1 

rH CO o i>- o w 

i-< -^ »o -^ o «-J 



CO <N O OJ 
00 Oi ^ »-• 
^ rH d C^ 






'3 



8 

■ 

I 



1 



I 

a 

9 



a 

t 
a 

OQ 

3 



c 

& 

d 



a 

3 ? 

'** 9 

© P 
«© 
■ c 






E" 

a: 






&C^ 



CO 

s 






CO CO S 
CO CO CO 



gl>- O 00 «5 o 

CO 1-H CO <o 1-) 

c<t M CO ci c^ CO 

00 «-H ^ ^<B tj< tj< 

f* 00 00 00 d 00 






CO ?0 00 O CO o 
1-1 *-•*-• c<i 61 CO 



o 



o o o o o o a» 
55 ^^. J^. >^ ^ ^ Q 



g^«3 
Tq>.§ 

Cg.Sc 



i-: 

Spqoo 



394 






2 









<> 
^< 

4 •J 

So 



•0»d ■» »»wid«ni J 



t>» CO 

• • • 

o o o 



o 



o 



10«Nw»PHoiqO 



onittiUO 



I *9 

9-^ 1-i 



O O CD w 



2S 

o o 



oojin^voijQ 



a 

< 

o 

s 

n 



PP» ORl 



^ ^ S 
CO CO ^ 

• • • 

o o o 



odd 



N xnoj. 



A%xA%a agioodg 



»0 CO 

cCco" 
o o 



3 

o 

CD 

O 

• 

o 

CO 

1—1 

d 

CO 

o 



o o 



O) 

• 

o 

CO 

o 

CO 



28 

• m 

o o 

• • 

o o 



-a 



o 



I 



ouin|OjV 






i~* i~* ^ 



lO o 



o 



CO 1-I 




8 

as 

•§1 



10«N 
8« eapuoiqo 





6 6^^^^ 
« o o a> 0) 0) 

QQQQQQ 


• 

1 


• • • • 

o o o o 



s 



ooi^vuO I 



oaiupvuQ 



il 



pttMioun : ^ 

o 



« 

r 

E 

I 



00 o ^ 

• • • 

r>- »-^ »-^ 

O 00 ^ 

• ■ • 

o -^ -^ 



CO 

d 



^ 
« 



00 

d 



• • 

Ci CO 

•H O 

28 

O O 

• • 



o 
c 



I ^ I 



^ 

PU 



a)9)oad- 



uo^ 



NPnox 



8P!i<* rnox 



S Oi ^ CO ^ 



I 



00 



O 

I 

5 

JS 

5 
S 



9 

-a 

d d 

Is 

«t I 

©•*; 
d o 

l| 

jafi 



.28 

1 it. 



0;5 



I 







i SSS 



25 '« 2 

0> »w^ ^* 

<N qa (N 



I 






1^ Cl 
C^ O) 






CD 2 "^ 

l^ "S ^ 

CO 



CD r* 

•^ CO 




B7 



s s s 



o 



c5 ^ 



s 3 9 3 <rs 



c^ u c^ 

V O V 



o 

V 






395 



396 Non- Protein Nitrogenous Compounds of Blood 

activity of the kidney with a corresponding retention in the blood. 
As will be noted, the figure for the uric acid just previous to death, 
27 mgm., is very high, even exceeding those reported for Case 1. 

Case 3 was found at autopsy to be typically of the interstitial 
type, although, as distinguished from Case 2, there was a marked 
accumulation of fluid during the last two weeks of life which en- 
abled us to secure large samples of pleural, ascitic, and subcutane- 
ous fluids for analysis. Uremic sjrmptoms did not show, however, 
until just preceding death, and it should further be noted that 
there was not a very marked accumulation of waste products 
until shortly before the fatal termination. Although the urea and 
uric acid reached a fairly high level, the figure of 5.3 for the 
creatinine is not high when compared with the other cases. 

Marshall and Davis'" have pointed out that urea is quite uni- 
formly distributed throughout the body. This is certainly true 
here with regard to the body fluids, as shown in the tabulated data 
on this case. A similar state of equable distribution throughout 
the body fluids and tissues was observed for creatinine. The 
same may be said in general for uric acid, if we except the spinal 
fluid, the difference here recorded being of interest in view of the 
position of relative isolation held by this fluid.'' 

DISCUSSION. 

The points of special interest in the tabulated data below 
bear brief discussion. Very high concentrations of uric acid 
and creatinine may be encountered in nephritis with retention of 
nitrogen. Uric acid has been found as high as 27.0 mgm., and 
creatinine as high as 33.3 mgm. per 100 cc. of blood. 

That the retention of the well known end-products of protein 
metalx)lism are not in themselves the cause of uremic symptoms 
would appear evident from the case of mercury bichloride poison- 
ing, in which uremic symptoms did not appear until more than a 
week after maximum concentrations of these substances in general 
had been attained. The possibility that a decomposition prod- 
uct of creatinine, such as methylguanidine, might play a part 

'•^ E. K. Marshall, Jr., and D. M. Davis: this Journaly xviii, p. 53, 1914. 
" See papers by H. Gushing, L. H. Weed, and P. Wegefarth: Jour. Med. 
Research y xxxi, pp. 1-176, 1914. 



V. C. Myers and M. S. Fine 



397 



II 

a ^ 



PfOVOUfl 



Nl*K)X 



8 II 



1-1 c^ 



i[- 



5 



Ok 



5a 

CO CO 






CD vH 

m m 

• B • 

o o o 

0) O V 

QQQ 



oapvuQ 



g or^ocor^&oo)cococ4<<4<o>oor^ 



9 






oxnni^vaiQ 



^ c«i-<cOcoc«ot>;Oeoo'^oocooooo 



pKmcrTJXi 



<i) I 



o 



8 I 



< 
D 



I 

I 8 



g OC^O^COiOb-Ot^O'^OOOO 
§ i-J'^QOiOi-iOOQdood'^OOOQOC^ 






t 

M 

o 

s 

•a 
B 



I 






E t^Sf3SSSSSSSS 9^^9 ! 



ooooooooo 



o o o o 



I 



I . 



o 



D 

h 

Q 
O 

n 



N mox 



c« ci .-^ o o o d o 



> 
o 



o 
o 

R 

a 

I 

8 



t 

a 



a 



e^ 



BP!i<*mox 



. 00 o 00 r^ t"* -^ « 
6 

«t l>.»OCOC0C0C0«-< 



• w* as 



I 



o 

9 



o 
'3 



3 



3 



o 



o 

o 



99 

3 



11 

8-3 g 
l|i 

ill 



-3 T3 2 ^ -d 2 -3 

"^ 3 2r •'^ 3 Zk C ^^ w M* fli • ** _• c ., «j"T" 



M 

•" o o 

• o s •• o 

:iU§|l 

: ««-■ ec c G ^ 



Ml 



8^ 



2 S ^ 2 2 

« « d d c 

« « § 5 § 

Q a i-s i-s "-3 



C5 



eS 

•-3 



7* S c M > Z B'Ti C 
s <D o ^ A S^ <$ ? 



TBI lOUBNAL OF BIOLOOICAL CHEUiaTRT, VOL. XIC, NO. 3 



a 

< 

m 

PS 







0ap«9JQ 



wwpuoiqo 



s 



r 
r 



r 






8 



00 






est 






•PHos imox 






s 

esi 



SI 






O 00 c* 

1^ CO •© 

o a r« 

O CO ^' 

o r« o 

• « • 

00 to ^ 



CO 



CO 



Ok u» Ok t^ 

00 o t^' «o 

00 a «o ^ 

^ CO CO CO 

o io e« «o 

• • • • 

00 ^ e>9 CO 



el 


CO 


M 


** « 


'•I 


o 


iO 


^ « 


oo 


e 


o 


^ o 


t^ 


o 


00 


00 ^ 

V4 


o 


o 


oo 


o •* 


r<. 


r* 


<D 


• • 



c* 



sg 



00 



s 



s*^ 



CO 



8r« t^ e« ^ 
«o t^ r« Ok 



S S £4 

S to «o 



3S 



O 00 
C« CO 



CO CO 
•^ r- 1-1 ^ 

>» ^ >» ^ 

flS «S «S CS 



S 



^ 3S 



CO CO ^ o 

Ok ^ eii o 



J5 



S 



CO 

00 

o 

04 



CO 

> 
o 



■a 
S *s S « 

* 3 kO lO 

N « • 

— "3 



Cl CS| 



o o 

^ O- od i>^ 

•X CQ ^H «ii< 



« r« e« S 

*^ e« "^ 

a c a •g 

•-» "n «-» ^ 



s 


s 


co 




s 


c« 


e« 


M 


C4 


M 


^ 


a 


CO 


C« 


O 


r«» 


<D 


u» 


•* 


•O 



• • • ^ u 

a c c -S S 
•n ■-» "n * « 






H 

fl 

m 



m 
o 

<4 




d 

c 






S B 

o a «> 



;^ 






a 
o 

eS 

a 



3 = 

•s .a 




M 



• 
•A 



OQ 



ki a E 



C oj • 



.9 



5 

a 



S 



•o 



^ 
J 



9 ^ 
o »- C9 

• -^ "c 
M CI. g 

* • £ 

> C3 3 



:;> 



s 



Q 



398 




suss 



-"SSSS l! = 



I a I 



Its, s I = 



f-3;l1 |ii J-3-i 



i-; 



1 1 ., Is 



III I -3 1 III III 



a 

5 







H (» PL4 t> 



• ■ 

S 













onnxT^WiQ 



a 

< 
as 

O 

S 

n 



pijw oun 



N«jn 



aio)CMd-ao|i4 



n wpiiopio 



L.. 

i BpnoBi«ioj. 



8 

s 









I 

e 
I 

I 






e« 



to 



CO 



CO 



00 «o 

• * 

»* 00 



e>9 CO ^ 



r<. r<. 



Ok CD^OOOOk M3 M» ^Oa 



e 



6 

I 



6 



s 






^ 






CO 



Q. 
o 



CO 



s 






M» o «D a o ei 

o ei t^ ci ^ c* 

««• e« Q 

*^ •-« w 

r« o 1^ «p 

M CO C« ^ 



:g3 

o o 



M e« 



CO 



c« 



*4 c« e* 



CO 



ss 



e4 



CO 

d 



& 



a 



s s s 



C« CO 



CI 

00 

csi 



ei « « «-• 



ci 

00 



00 e 
^ c« 

csi CO 



00 2 w "> r* 

o o a> 

c3 a 

3 3 



» C ^ 1^ 



c« 

C8 





s 




ss 


as 


Dee. 
Doc. 






30 




.9 



0) 



II 



i< li^ 



.a 






3 
o 

O 






S 



C4 



CSI (D 






eo 






ti 



ao 






oo 






e> o ^ CI 

^ ^ CI C4 



c 



s 



CI 



CI 



400 



V. C. Myers and M. S. Fine 



401 



in the development of uremic symptoms needs to be considered, 
although there are as yet no definite data to support this view.^* 
In six cases of gout examined in this laboratory, uric acid in 
100 cc. of blood ranged from 3.8 to 5.8 mgm. The far higher 
values for uric acid in the above cases of uremia, in which gouty 
symptoms were absent, ar-e of interest in view of the controversy 
existing with regard to the relation between the retention of uric 
acid in gout and the clinical symptoms.^' 

TABLE I. 

Summary of dcUa on uric acid, creatinine, and creatine in the blood. 



CABE 



1 
2 
3 
4 
5 

6 

7 

8 

9 

10 

11 

12 

14 

15 

17 



CLINICAL DIAQNOSIS 






S 



% 

H 



H 

Z 

M 

< 

H 

PS 
o 



Per 100 CO. 



I 



Hg poisoning, uremia 

Chronic interstitial nephritis, uremia 

Chronic interstitial nephritis, uremia, edema. . . 

Chronic interstitial nephritis, uremia 

Chronic interstitial and parenchymatous nephri- 
tis, uremia 

Nephritis of mixed type, uremia 

Interstitial nephritis, uremia 

Interstitial and vascular nephritis, uremia 

Acute nephritis of mixed type, uremia, edema 

Interstitial and vascular nephritis 

Chronic interstitial nephritis, edema 

Chronic vascular and tubular nephritis 

Chronic interstitial nephritis 

Chronic parenchymatous nephritis 

Chronic interstitial nephritis, aortic insufficiency 



mgtn. 

15.0 
27.0 
15.4 
11.4 

14.0 
13.4 
8.0 
9.1 
5.5 
5.5 
4.0 
4.5 
4.1 
4.2 
6.3 



mgm. 

33.3 

20.0 

5.3 

16.6 

14.7 
7.4 
4.8 

11.0 
7.0 
4.6 
2.1 
1.8 
3.3 
1.5 
2.1 



mgm, 

19.4 
31.4 
21.3 



15.2 
13.1 



Another point which appears to possess some little significance 
is the increase in the creatine content of the blood in those cases 
showing a very marked rise in the uric acid. There was not a 

" In this connection the papers by N. E. Ditman and W. H. Welker : New 
York Med. Jour., Ixxxix, pp. 1000, 1046, 1091, 1134, 1909, are worthy of note. 
Abo W. M. Kraus: Arch. Int. Med., xi, p. 613, 1913. 

" See M. S. Fine and A. F. Chace: Jour. Pharm. and Exper.' Therap., 
vi, p. 219, 1914; Jour. Am. Med. Assn., Ixiii, p. 945, 1914. 



402 Non-Protein Nitrogenous Compounds of Blood 

complete retention of uric acid, but even if no uric acid were 
eliminated during the last days of life, the amount of retention, 
if anything, exceeded the probable endogenous formation (Cases 
2 and 3). From the data presented on Cases 3 and 25, we may 
assume that uric acid is quite evenly distributed throughout the 
fluids and tissues. Since the increased creatine is suggestive of 
increased tissue destruction," it is possible that this in part ex- 
plains the very high uric acid. 

The distribution of these products in the spinal fluid is worthy 
of note; whereas urea and creatinine approach the concentrations 
found in the other fluids in uremia, the creatine and uric acid 
are very low or practically absent. 

The comparison between the t'etention figures for uric acid 
and creatinine is of interest. In some cases the creatinine is 
decidedly higher than the uric acid, while in others the reverse 
is true (Table I). *In the majority of the cases a high creatinine 
is found to accompany a very high urea, while in those cases in 
which the creatinine is low the urea is generally low. There 
appears to exist some parallelism, then, between the accumu- 
lation of urea and creatinine, and, as pointed out above, also be- 
tween the uric acid and creatine, although it has not been possible 
to correlate these parallelisms with any specific typos of nephritis. 

'* Myers and Fine: this Journal, xv, p. 283, 1913. 



STUDIES ON A METHOD FOR THE QUANTITATIVE 
ESTIMATION OF CERTAIN GROUPS IN 

PHOSPHOUPINS. 

By MARY LOUISE FOSTER. 

{From the Hull Laboratories of the University of Chicago, 

Chicago.) 

4 

(Received for publication, January 11, 1915.) 

In 1903 Baskoflf^ published a method for the quantitative 
estimation of glycerol and the nitrogen complex. He dissolved 
lecithin in absolut<3 alcohol and saturated this with hydrochloric 
acid gas. The solution was heated with a reflux condenser until 
hydrolysis and esterification of the fatty acids had taken place. 
The alcohol was then almost entirely removed by heating, the 
solution dilute<l with water, and the esters of the fatty acids were 
shaken out with ether. The ^solution at this stage contained the 
nitrogen complex, glycerophosphoric acid, free phosphoric acid, and 
glycerol. A portion of this was treated according to the method 
of Stanek-KieseP with potassium triiodide, and the nitrogen of 
the precipitate was determined by Kjeldahl. It gave 1.09 per 
cent, which represents 11.07 per cent choline. A second portion 
was evaporated to dryness and extracted with absolute alcohol 
to remove the choline hydrochloride. The residue was treated 
with water and again taken down to dryness, a process repeated 
several times. It was finally transferred to a Fanto-ZeiseP 
apparatus where it was concentrated to 5 cc, 15 ec. of hydriodic 
acid were added, and it was heated to boiling. At the same time 
a stream of CO2 was passed through the whole apparatus. The 
vapor of isopropyl iodide which escaped from the upright cooler 
was led for purification from hydriodic acid vapor through a 
wash bottle containing rod phosphorus under water, and finally 

' A. Baskoflf: Ztschr. f. physiol. Chem.^ Ivii, p. 4.35, 1908. 

« V. Stanek: ibid,, xlvi. p. 283, 1905. 

« S. Zeisel and R. Fanto: Ztschr. f. anal. Chnn., xlii, p. 549, 1903. 

403 



404 Quantitative Estimation of Phospholipins 

into an alcoholic solution of silver nitrate. The silver iodide 
was weighed, and by calculation gave the amount of glycerol. 
Baskoff found in lecithin 10.61 per cent, calculated as CsHsCOH)!,- 
the theoretical amount is 11.8 per cent as CsHsCOHjs, 5.2 per 
cent as CsHs, on Trier's* formula. 

The above work suggested the possibility of sphtting off the 
glycerol group and the methyl groups attached to the nitrogen 
in one operation, but at different temperatures, by means of hydri- 
odic acid. Preparations of phosphoUpins, i.e., lecithin, the alcohol- 
soluble compound, and kephalin, the alcohol-insoluble compoimd, 
were made according to Waldemar Koch's* method, except that 
care was taken to use only anhydrous solvents. By this method, 
after repeated purification, the alcohol-soluble compound gave 
by Arnold-Gunning Kjeldahl for N = 1.91 per cent, and by the 
Neumann method for P = 3.05 per cent weighed as MgiPjOy; 
the alcohol-insoluble preparation gave N = 1.89 per cent; P = 
3.73 per cent. 

The following table shows the results obtained by various 
investigators of these substances: 

Kephalin per cent per cent 

Thudichum 1 .68 4.27 

W. Koch* 1.78 3.84 

Neubauer« 1.65 3.45 

My preparation 1 .91 3 .05 

Lecithin 

Diakonow^ 1.8 3.8 

Thierfelder» 2.08 3.97 

McLean* 1 .88 3.95 

Eppler 2.09 3.95 

Thudichum 2.03 4.29 

W. Koch 1 .80 3 .79 

Erlandsen'® 1 .87 3.95 

My preparation 1 .89 3 .73 

* G. Trier: Ueber einfache Pflanzenbasen und ihre Beziehungen turn 
Aufbau der Eiweiaastofe und Lecithinef Berlin, 1912. 

» W. Koch: Am. Jour. Physiol, xi, p. 319, 1904. 

•E. Neubaucr: Biochem. Ztachr., xxi, p. 321, 1909. 

'Diakonow: Med. chem. Untersuch.f 1867, p. 21. 

•H. Thierfelder: Ztschr. f. physiol. Chem., xlvi, p. 518, 1905. 

• H. McLean: ibid., Ivii, p. 297, 1908. 
" A. Erlandsen : ibid., li, p. 71, 1907. 



M. L. Foster 



405 



While these figures vary considerably among themselves, the 
ratio 1 : 1 between N and P is fairly well maintained and shows 
that these phospholipins, although obtained by sundry methods 
of extraction, nevertheless belong to the class of mono-amino- 
mono-phosphatides, as stated by Erlandsen. 

Before applying the method for determining the glycerol m 
these preparations, controls were run with pure glycerol. The 
glycerol (Merck's preparation) was so diluted that 0.5 cc. con- 
tained 0.2 to 0.3 gram of glycerol. This was placed in a small 
round-bottomed flask with 1 gram of ammonium iodide and 
15 cc. of hydriodic acid, and connected with an upright 15 inch 
condenser having a thermometer and a side tube, leading to a 
Geissler bulb which contained 2 per cent solutions of sodium 
carbonate and potassium arscnite, as recommended by Klinger 
and Kreutz." The flask containing the glycerol was heated in a 
glycerine bath to 112° to 113°, at which temperature a cloud 
appeared m the silver nitrate solution. The heating was con- 
tinued at this temperature until the precipitate became crystal- 
line and settled, leaving the supernatant Uquid perfectly clear. 
The reaction probably takes place according to the equations: 

CH2OHCHOHCH2OH + 3HI = CH2ICHICHJ + 3H2O 
CH2ICHICH2I + 2HI = (CH,)2CHI + 2I2. 



AMOUNT OF 
OLTCEROL UBED 


AMOUNT or Agl 

OBTAINED 

gm. 

0.6323 
0.6296 
0.6167 
0.1800 


AMOUNT or OLTCEROL 
FOUND 


PER CENT 

• 


gm. 

0.2466 
0.2292 
0.2306 
0.0719 


gm. 

0.2479 
0.2077 
0.2026 
0.0704 


100.4 
90.62 
87.8 
98.4 



The time required for completion of the reaction was two to 
three hours, the temperature remaining constant at 112° to 113°. 
The temperature was finally allowed to rise to 120°, at which 
point if no further precipitation took place, the process was dis- 
continued. The weight of silver iodide multiplied by the factor 



" H. Klinger and A. Kreutz: Ann. d. Chem., ccxlix, p. 147, 1888. 



4o6 Quantitative Estimation of Phospholipins 

0.3922 gives the weight of glycerol found. This method was a 
combination of the Fanto-Zeisel and Hewitt and Moore*- methods. 

A blank gave negative results and warranted the conclusion that 
the Gcissler bulb was holding back all hydriodic acid and iodine. 

The phospholipins were then tested in the apparatus. Keph- 
alin was tested first and gave a precipitate at 112°, corresponding 
in every detail to that obtained with glycerol. Even on heating 
a second time no turbidity appeared in the silver nitrate solution. 

KEPHALIN WT. OP QLTCEROL 
NO. C8ED Agl. CALCULATED GLYCEROL 

gm. gm. gm. percent 

1.0.2181 0.0507 = 0.0199 = 9.11 

2. 0.7260 0.1517 = 0.a595 = 8.21 

3. 0.3242 0.0628 = 0.0246 = 7.62 

On raising the temperature no further precipitation in a fresh 
silver nitrate solution took place. This indicates that no methyl 
group was liberated. 

NO. gm. gm. gm. percent 

1. 0.2438 lecithin 0.0563 Agl at 112** = 0.0220 glycerol = O.Oo 
1 a. 0.2438 lecithin 0.0368 Agl at 180'' = 0.00235 methyl = 0.96 

2. 0.2235 lecithin 0.0548 Agl at 112** = 0.0215 glycerol = 9.61 

3. 0.2615 lecithin 0.0668 Agl at 112'' = 0.020!? glycerol = 10.01 

4. 0.4972 hydrolyzed 

lecithin 0.1214 Agl at 112** = 0.0476 glycerol = 9.57 

5. 0.4972 hydrolyzed 

lecithin 0.1135 Agl at 112° = 0.0445 glycerol = 8.95 

6. 0.4972 hydrolyzed 

lecithin 0.1182 Agl at 112** = 0.0464 glycerol = 9.32 

Average for Nos. 1, 2, and 3, unhydrolyzcd = 9.55 per cent; 
for 4, 5, and 6, hydrolyzed according to Baskoff's method = 9.38 
per cent. It is evident that this process is unnecessary, the hy- 
driodic acid alone being sufficient. 

In ever>'^ instance after apparently complete precipitation of 
iodide from the glycerol had taken place, the temperature was 
run up slowly to 300°, but no further clouding of the fresh silver 
nitrate solution took place except in 1 a, where a precipitate was 
obtained. This was, however, too much to be wholly glycerol 
and too little for three methyl groups. Two facts had become 
evident: the glycerol group in the phospholipins under consider- 

»« J. T. Hewitt and T. S. Moore: Jour. Chcin. Soc, Ixxxi, p. 320, 1902. 



M, L. Foster 407 

ation is very easily and almost quantitatively split off by hydri- 
odic acid; and, second, either the method is ineffective where 
imide methyl groups are concerned; or there is no methyl in the 
substance. The first fact will necessitate the recalculation of 
Waldemar Koch's results. In the case of kephalin: 

0.3488 gm. substance; 0.0945 gm.AgI = 0.0378 gm. glycerol= 10.8 per cent, 
instead of 
0.3488 gm. substance; 0.0945 gin. Agl = 0.006036 gm. CIIs = 1.73 per cent. 

This admits of no methyl group in kephalin and agrees with 
Winterstein's^^ finding of 10.2 per cent glycerol, or 1.63 per cent 
CHa, as given by this method. In the case of egg lecithin Koch 
got a precipitate of 0.0760 gram of silver iodide from 0.320 gram 
of the substance below 240°. Calculating this as glycerol, the 
yield is 9.5 per cent, and leaves for methyl only 4.3 per cent 
instead of 5.80 per cent, or two methyl groups instead of three. 

These irregularities and inconsistencies suggested the necessity 
of testing choline, which has so often been called the alkaloidal 
base of lecithin. Choline chloride was prepared by the method 
of Wurtz,^^ as modified by Renshaw.^^ 

Attempts were made to determine the methyl groups in this 
preparation by the use of the Fanto-Zeisel method given above, 
but no adequate cleavage of the methyl group took place. Accord- 
ing to Decker,*® dry heating is necessary and this is the chief 
advantage of the double barrel distilling flask used in the Herzig 
and Meyer method. By using separate cups filled with paraffin 
the temperature can be absolutely controlled, as is not the case 
in the sand bath, where the temperature about the immersed 
flask was found to varj'^ 40° to 50° in spots a few inches apart. 
The tube connecting the two flasks was bent into an inverted U, 
so that the flasks were immersed nearly to the stoppers in paraffin. 
The temperature was allowed to rise slowly to 140°, where it 
was held for half an hour, then allowed to rise again to 180° to 

'• E. Winterstein and O. Hicstand: Ztschr.f. phyaiol. Chem.y liv, p. 300, 
1907-08. 

'* A. Wurtz: Ann. d. Chem.y Supplement, vi, p. 200, 1868. 
'* K. R. Renshaw: Jour. Am. Chem. Soc, xxxii, p. 128, 1910. 
'• H. Decker: Ber. d. dciUsch. chem. Gesellsch., xxxvi, p. 2895, 1903. 



4o8 Quantitative Estimation of Phospholipins 

185°, at which point a precipitate was usually formed in the 
silver nitrate solution. Here the temperature was held until 
the reaction was over, a process lasting four to five hours. The 
precipitate was then treated as with glycerol. 

The reaction takes place in two stages. According to Wurtz, 
the change of the alkyl groups of choUne into the iodide form 
occurs between 120° to 150°. The boiling point of hydriodic 
acid is 127°, slightly raised in this instance by the presence of 
the ammonium iodide. 

N(CH,),CH2CH20HC1 + 2HI = C2H4lN(CH,),I + HjO + HCl 

This reaction is slow and requires time. It seems probable 
that it is because this reaction has been incomplete that repeated 
successive heatings are necessary. Xo way of knowing when 
this reaction is complete has been discovered, however. At a 
higher temperature dissociation takes place, and methyl iodide, 
which decomposes less readily than ethyl iodide, is split off, 
and passes over into the alcoholic silver nitrate, where it is 
precipitated. 

Besides the substitution of a paraffin bath for the sand bath, 
some other modifications were introduced: a current of water 
heated to 40° to 50° ran through the condenser while the Geissler 
bulb containing 2 per cent solutions of sodium carbonate and 
potassium arsenite was inmiersed in water heated to 50° to 60°; 
a second flask containing the arsenite solution, also heated, was 
introduced as a precautionary measure; and finally the gases 
were bubbled through a tall cyUnder of alcoholic silver nitrate 
by means of a modified Folin tube. This effected a more com- 
plete precipitation. A second, third, and even fourth heating 
increased the yield, although the evidence of the settling of the 
precipitate in each previous heating had been taken as sufficient 
proof that no more iodide was forming. This reheating was made 
by turning back into the first bulb the hydriodic acid which had 
collected in the flask connected directly with the condenser, and 
then repeating the process exactly as in the first instance. Much 
more uniform results were obtained after the slow digestion was 
introduced, but a white smoke which appeared as a constant 
factor was not satisfactorily explained. 



M. L. Foster 



409 



NO. 



1 

2 
3 
4 
5 
6 
7 
8 



■ AMOUNT OF 
CHOLINE CHLO- 
BIDB 



gtn. 

0.1000 
0.0431 
0.0762 
0.0418 
0.0311 
0.0416 
0.0401 
0.0655 



AMOUNT OF 

Agl ON riBST 

BBATINO 



fffll- 

0.4286 
0.1575 
0.2907 
0.1805 
0.1451 
0.1964 
0.1895 
0.3075 



AMOUNT OF ; r-TT »xi> ••a.* 
CHtFOB riBST CHlFOB riBOT 

--.~« HEATING 

HEATING 



gm. 

0.0274 

0.01006 

0.01857 

0.01153 

0.00927 

0.01255 

0.012105 

.01964 



per cent 

27A 

23.3 

24.4 

27.87 

29.8 

30.01 

30.18 

29.68 



TOTAL CHi 

per cent 

35.1 

46.3 

39.3 

32.27 

37.3 

42.6 

40.01 

36.99 



The slow digestion first used on No. 4 brought about fairly 
uniform results for the first yield, but although applied also in the 
succeeding heatings the final yield shows considerable variation. 
The factor 0.06388 was used to convert the silver iodide into 
methyl which should amount theoretically to 32.2 per cent of 
choline chloride. If the hydroxyethyl group is also liberated and 
calculated as methyl, the total per cent of methyl would be 42.9 
per cent. 

A mixture of glycerol and choline chloride was then tried. The 
separation of glycerol was sharp and complete at 112° to 120°, 
no more precipitate appearing in the silver nitrate solution as the 
temperature was raised. The second precipitate began to appear 
usually at 175° and was most abundant at 185°. 



NO. 



3 



AMOUNT 
UBED 



gm. 

Glycerol 0.1620 

Choline CI.... 0.1682 

Glycerol 0.0719 

Choline CI.... 0.1000 

Glycerol 0.0719 

Choline (1... . 0.1322 



TEMPEBA- ; WEIGHT OP 
TUBE Acl 



C. 

112-120** 

180^ 

112-120** 

183^ 

112-120^ 

160-243^ 



CALCULATED 
TO 



gm. i gm. [percent 

0.3811 I 0.1495 Glycerol . 92.2 

0.5607 I 0.0359 CHs 21.3 

0.1882 0.0738 Glyceror 102.6 

0.5486 0.0351 CH3 35.1 

. 1782 , . 0698 Glycerol 97 . 2 

0.8927 0.0562 CHs 44.2 



These figures represent the total yield obtained after seveial 
heatings. 

With the improved apparatus analyses of the phospholipins 
were repeated. 



4IO Quantitative Estimation of Phospholipins 



NO. 



Lecithin. 



2 I Lecithin. 



3 Egg lecithin. 



AMOUNT 
USED 



TEMPERA- 
TUBE 



gm. 
0.3642 

0.4927 

0.2945 



4 i Lecithin HCl . ' 0.2868 



WBIOHT OF 1 
Agl 



CALCULATKD 
TO 



C. 
112° 

180-185^! 

112° ' 

160-240°! 

112° 

180-190° 

112° [ 

180-190°] 

5 I Kephalin ' 0.1613 ' 112° 

i ' 180° ' 

6 I Lecithin HCl.! 0.1513 \ 112° 

I ' I 180-190°' 



gm. 

0.1012 
0.0766 
0.1238 
0.1122 
0.0658 

0.0850 
0.2743 
0.0338 



gm. 

0.0397 

0.0049 

0.0496 

,0.0072 

0.0258 

10.1065 

0.0333 

0.01752 

0.01326 



Glycerol 

CHs 
Glycerol 

CHs 
Glycerol 

CHs 
Glycerol 

CHa 
Glycerol 



0.039 0.1529 Glycerol 
0.3474 0.02219 CH3 



percent 

10.9 
1.34 

10.05 
1.45 
8.75 
3.62 

11.6 
6.10 
8.21 

9.96 
15.1 



These results for glycerol are slightly higher than those obtained 
by the use of the simple apparatus. If any reliance is to be 
placed on this method for the determination of the methyl groups, 
it is evident that lecithin prepared from sheep's brains by the use 
of acetone, alcohol, and ether contains only one methyl group, 
the ratio of N: P: CH8 = 1.89: 3.73: 1.45 = 0.12: 0.09. Koch found, 
allowing 9.5 per cent glycerol as indicated by the precipitate 
below 240°, 3.80 per cent for CH3, or a ratio of N : P : CH3 = 
1 .80 : 3.79: 3.80 =1:1:2. The egg lecithin gives a slightly lower 
figure for glycerol and a much higher result for methyl, while 
the lecithin hydrochloride prepared by treatment of the above 
lecithin with cadmium chloride and removal of the cadmium with 
hydrogen sulphide as prescribed by Thudichum gives still higher 
results. The result would indicate for this preparation three 
methyl groups. Kephalin gave no precipitate at the higher 
temperature, indicating no methyl groups. A comparison of 
Xos. 4 and 6 shows widely varying results for samples from the 
same bottle. This is probably due to the different amounts of 
time for which each was tested; No. 4 was given two heatings 
lasting about seven hours, while No. 6 was given three heatings 
amounting to ^bout ten hours in all. Neither was heated long 
enough to give final negative results. It is evident that pro- 
longed heating introduces errors — a fact already noticed by 
Stritar — due to the fonnation of various alkyl iodides by the 
hvdriodic acid. 



M, L. Foster 411 

These varying results suggested the necessity of testing the 
method on some of the constituents of lecithin other than choline. 

SUBSTANCE 'SSSS' TIB^ED '0"^*^= CAI^ULATEDI 

Om. gm. pa cent per cent 

Palmitic acid . 1198 .010 at 190^ 0.5 CHa 

Oleicacid 0.3194 0.0137 " 190*» 0.2 CHs 

Mono-ethyl amine hydro- 
chloride 0.1149 0.0939 9.7 CaH* 35.6 CizH» 

Trimethyl amine 0.0513 0.432 53.79 CH, 47. 17 CHs 

A sixth heating of the trimethyl amine hydrochloride gave 
negative results in the silver nitrat<; solution. White smoke was 
a constant accompaniment of the experiment except in the sixth 
heating, where there was none. Smoke and precipitate^ began to 
appear at 150°, a much lower temperature than in the analysis of 
choline chloride or ethyl amine hydrochloride. The total time 
of heating was about seventeen hours. It is difficult to account 
for this excessive yield. 

Some methyl iodide was repeatedly fractionated and a portion 
boihng between 42° to 43° was introduced into a stoppered bottle 
and dropped into the bulb of the apparatus containing hydriodic 
acid and given the usual treatment. The precipitate appeared 
in the silver nitrate solution almost immediately, the stopper 
having been forced out by expansion. The temperature was, 
however, raised to 180° and the heating continued till the solution 
had become clear. 0.2057 gram of methyl iodide yielded 0.3229 
gram of silver iodide = 94.84 per cent of the total. Some methyl 
iodide in a similar stoppered bottle was dropped directly into 
some alcoholic silver nitrate solution. 0.1505 gram of methyl 
iodide yielded 0.2440 gram of silver iodide = 97.97 per cent. 
The analytical method with the apparatus gives then a yield of 
96.7 per cent. It is evident from this experiment that the methyl 
iodide is not lost in transit, but comes over nearly quantitatively, 
if the direct precipitation be taken as the standard. The only 
jxjssible explanation for the high yield in the experiment with 
trimethyl amine hydrochloride would seem to be that hydriodic 
acid forms unsaturated compounds which yield more than one 
methyl iodide equivalent. Work by Dr. Xef suggests this as 
possible and probable. 



412 Quantitative Estimation of Pliospholipins 

From these results it is evident that the method in its present 
shape is not applicable to the analysis of phospholipins for the 
purpose of determining their formulae. Hydriodic acid is a 
powerful reducing agent and attacks both methyl and ethyl 
groups. Kahan^^ found that methyl iodide heated to 265° for 
three hours gave no gas, but heated to 270® it went over into 
methane and ethylene. At 185° methyl iodide is fairly stable, 
giving only slight evidence of gas. This is not true of ethyl 
iodide, which goes over into ethane, ethylene, and hydrogen. 
I found that the ethyl amine hydrochloride gave the sublimate 
and smoke so characteristic of all the choline chloride determi- 
nations. The gas or smoke was collected and found to be an 
unsaturated hydrocarbon. Goldschmiedt^' has studied many 
compounds containing the methyl group in various positions, 
using this method. He finds that the sphtting off of the alkyl 
group is dependent upon the structure of the compound, the 
nature and the position of the groups, and that it is impossible 
quantitatively to differentiate N-CHs and OCH3. The methyl 
group often seems to wander from nitrogen to oxygen. It is 
possible that longer digestion below the boiling point of hydriodic 
acid may result in a wider separation by temperature of the methyl 
and ethyl iodide and admit of a fractional separation of these two 
alkyl groups. 

SUMMARY. 

1. By the use of a paraffin bath, the temperature of which can 
be controlled, it is possible to obtain by a modified Herzig and 
Meyer method a sharp separation of the glycerol and alkyl groups 
in certain phospholipins, the former reacting at 112° and the 
latter at 180°-190°C. 

2. The glycerol is obtained in almost theoretical quantity. 

3. The ethyl and methyl groups are not so quantitatively 
determined and differentiated. 

4. Thudichum's preparation of lecithin shows the three methyl 

'" Z. Kahan: Jour, Chem, Soc, xciii, pt. i, p. 132, 1908. 

'*0. Goldschmiedt and A. Kirpal: Monalschr. /. Chem.j xvii, p. 491, 
1890. G. Goldschmiedt: ibid.j xxvii, p. 840, 1906. G. Goldschmiedt and 
O. Honigschmid: ibid., xxiv, p. 681, 190.^. 



M. L. Foster 413 

groups commonly attributed to lecithin; but our preparations 
show less. 

5. The method shows no methyl or ethyl groups in kephalin.^* 

6. The analysis of synthetic choline chloride shows that hydri- 
odic acid does not spUt off the methyl groups quantitatively. 

7. Ethyl amine hydrochloride gives many of the phenomena, 
viz.y the white smoke and the sublimation in the U tube, noted 
in the course of the reduction of the choline chloride. These are 
not so marked in the case of trimcthyl amine hydrochloride. 

I wish to express here my thanks to Professor A. P. Mathews 
and to Professor F. C. Koch, to whom I am indebted for many 
helpful suggestions. 

Smith College, 
Northampton, Mass. 



^•MacArthur announces in the Jour. Atn. Chem. Soc, xxxvi, p. 2397, 
1914, that kephalin contains neither choline nor ncurine. 

THB JOURNAL OF BIOLOGICAL GHBMISTRT, VOL. XX, NO. 3 



THE INFLUENCE OF THE PLANE OF PROTEIN INTAKE 

ON GROWTH.! 

By E. V. McCOLLUM and MARGUERITE DAVIS. 

{From the Laboratory of Agricultural Chemistry of the University of Wis* 

consin, Madison.) 

(Received for publication, January 30, 1915.) 

The experience of animal husbandrymen has established the 
fact that young animals need a relatively high plane of protein 
intake in order to make the best possible rate of gain. McCoUum^ 
has shown with pigs that the per cent of the ingested nitrogen 
retained for growth when the ration contains above 10 per cent 
of protein is independent of the plane of protein intake. It 
follows that when the consumption of protein is high, the reten- 
tion is correspondingly high. In the same paper it was pointed 
out that although the character of the protein mixture supplied 
by the cereal grain is such that it can be utilized for growth only 
to the extent of 20 to 25 per cent, yet growth (retention of nitro- 
gen) continues to take place when the protein content of the 
ration drawn from wheat or oats was as low as 6.63 to 7 per cent. 
Lower protein levels were not fed. 

Osborne and MendeP have exhibited the curves of growth of 
rats fed "protem-free milk" with casein to the amount of 4 to 
31 per cent of the food mixture. These curves indicate thftt 
rats cannot maintain their body weight when less than 12 per cent 
of casein is fed. With 18 per cent of casein in the diet normal 
growth was secured, but rapid failure of the animals ensued when 
31 per cent of casein was contained in the diet. Their curves 
sliow, further, that with 4 per cent of edestin there was steady 

* Published with the permission of the Director of the Wisconsin Ex- 
periment Station. 

2 E. V. McCollum: this Journal, xix, p. 323, 1914. 

•T. B. Osborne and L. B. Mendel: Ztschr. /. physiol. Chem., Ixxx, pp. 
341-50, 1912. 

415 



4i6 Influence of Protein Intake on Growth 

decline, while with 6.5 per cent of this protein body weight was 
just maintained. With 9 per cent of edestin a small amount of 
growth was attained, somewhat better growth with 12 per cent, 
and normal growth with 18 per cent of edestin. They drew 
the conclusion that the lowest Umit of protein content which can 
lead to growth is 7 to 9 per cent of the food mixture. We have 
pointed out that these results are not in harmony with our own 
observations.* 

Since it is a matter of great importance to know what plane 
of protein intake from various sources is essential for body main- 
tenance, and the lowest plane at which the maximum rate of 
growth is secured, we have undertaken to establish these points 
with certainty for a variety of foodstuffs. 

In our experience even normal growth to the normal adult 
size and continued maintenance does not necessarily indicate 
perfect nutrition. Only when the animals reproduce and nourish 
their young normally, and repeat this at normal intervals, can 
it be said that the ration is physiologically sufficient. 

In the present paper we present curves showing the behavior 
with respect to growth of rats fed a mixture of skim milk powder 
(Merrill-Soule), dextrin, and butter-fat. The milk powder con- 
tent was varied so as to make the protein content of the rations 
YBxy between 2 and 10 per cent. For comparison we include 
also the curves of rats fed with wheat protein (6 per cent), wheat 
embryo protein (4 per cent), and egg protein (2.45 per cent). 

•In such experiments as those described it is of the greatest 
importance that individual records shall not be made the basis 
of comparison. A number sufficient to exclude the possibiUty 
of error, due to individual variations, should be employed. The 
curves which we present in this paper represent all our experi- 
ence with the rations described. The similarity of the perform- 
ance of all the animals of each group is convincing evidence that 
the results are conclusive. All the experiments were carried out 
at the same time and in the same room. 



* E. V. McCollum and M. Davis: Proc. Ain. Soc. Biol. Chemists, this 
Jotirnal, xiv, p. xl, 1913. 



E. V. McCollum and M. Davis 417 

SUMMARY OF RESULTS. 

1. It is evident from the curves shown that the lowest plane 
of protein intake derived from milk which can maintain young 
rats without loss of body weight is 3 per cent of the food mixture. 

2. There is a progressive increase in the rate of growth with 
rations derived fropi milk, as the plane of protein intake is raised 
between 3 and 8 per cent of the diet. 

3. It is evident that for a time at least rats may grow at about 
half the normal rate when the protein is supplied by the wheat 
kernel to the extent of 6 per cent of the food mixture. 

4. 2.45 per cent of protein derived from desiccated egg is not 
sufficient to maintain young rats without loss of body weight. 

5. During six weeks a ration carrying but 4 per cent of protein 
from wheat embryo compares favorably with a similar plane of 
protein intake derived from milk powder, apd is somewhat better 
than 6 per cent of protein from the entire kernel. 

6. We believe that- this plan of experimentation offers a valu- 
able method of comparison of the proteins from various sources, 
provided all deficiencies are made up by suitable additions. 



4i8 Influence of Protein Intake on Growth 




Chart I. (7 males.) Shows a steady loss of body weight in six out of 
seven of the animals on a diet containing but 2 per cent of milk protein. 
One has maintained its weight during 49 days on this very low plane of 
protein intake. 



The ration consisted of: 

per cent 

Milk powder 6.0 (34.0 % N X 6.25) 

Dextrin 81.0 

Butter-fat 5.0 

Agar-acar 2.0 

Halt mixture 6.0 



COMPOHITION or THE 8A.LT MIXTURE 

gm. 

NaCl 16.00 

Na citrate 3.70 

KjHPO* 34.22 

CaH4(I*0«); 0.89 

MgSO«(anhydrouB) 1 .90 

Mg citrate 7.00 

Ca lactate 57.02 

Fe citrate 2.00 



E. V. McCollum and M. Davis 



419 




Chart II. (2 males, 3 females.) Shows maintenance of body weight 
during 100 days with a ration containing 3 per cent of milk protein. 
The ration consisted of: 

percent 

Milk powder 9.0 

Dextrin 83.0 

Butter-fat 6.0 

Acar-acar 2.0 

Fe citrate 0.1 



420 Influence of Protein Intake on Growth 



fa 

ic 



5V> 



y-p 



y« 





S9 





wtmns 



Chart III. (5 males.) Shows the behavior of rats with a ration con- 
taining but 3 per cent of protein from milk. The mineral content of the 
ration in Chart II was derived solely from the 9 per cent of milk powder 
it contained, and was therefore very low. In order to show whether the 
low salt content or the protein content was the limiting factor in the ration 
of this group of animals a salt mixture was added. It is thus made apparent 
that the low protein content is the limiting factor. The rats behave just 
as did those without salt addition. 

The salt addition consisted of 5 gm. per 100 of ration of the salt mix- 
ture fed, the rats receiving 2 per cent of protein of ijiilk (Chart I). 



42 2 Influence of Protein Intake on Growth 




fig 
- -^ 

I « 

O 0) 

M O 

M O 

JS5 



5^ 



.3 

u 
08 

g 

O 
M 

00 

O 



V 



V 



08 










^ e 



h o8 

Ud o 

.■S II 

& » 

o. » 



-3 e o o o 

V OC iC 10 M 



"3 



I 



00 

OS 
g 



l-H « 



o o .. 

73 -^3 ^ 

•*^ is 'S 

7 00 

.2 o C 

es * « 

^ £ o 
g tc -^ 

•^ "S -^ 

S |H 

o &< 



c T S 

S O oc < 



^ >• 



2- Z- 




Chart V. (3 males, 2 femalei 
Dotted line » normal cunre. 
The ration consisted of: 



Milk powder...- 

Dextrin \ 

BaU«r-fAt \ 



Journal of Biological Chem 



E. V. McCollum and M. Davis 



423 



JLto 



tU 



174, 






r- 



fo 















/ 














V 


/ 


« 










'A 


/, 


'/ 







% 




A 


/ 


* 

/J 


/ 


— i 






y 





1 


7 




/ 




/ 


I- 


/ 




y 


— e 






'/ 


/ 




/ 





g 


^ 


V 


r 

/ 


r 




/ 


f 
/ 




V 


/ 




/ 


^ 
/ 


\ 




'/ 


/ 












1^ • •••••' 



Jo 

Chart VII. (3 males, 1 female.) Illustrates the growth of rats re- 
ceiving 8 per cent of milk protein. The growth is in all cases somewhat 
better than the normal expectation curve. Dotted curve = normal curve. 

The ration consisted of : 

Milk powder 24 . 

Dextrin W.O 

Bulter-tat 5.0 

Agar-agar 2.0 



424 Influence of Protein Intake on Growth 









■"S 


N 
















s 


N 




\ 


















■^ 


N, 


\ 


\, 


H 










• 


^ 


V 


\ 




\ 














s 


\ 


\ 


\ 


\ 






N 








\ 


^ 


\ 










•*.^ 


N 






\ 








V 


\, 






\ 


\ 




\ 










X 


V 




\ 


\ 














V 


\ 


» 


>5 


N^ 




1 






•k*V^ 






\ 








"^ 








\ 






""n 


^ 








• 


X. 








\ 






^ 


N 


• 
% 


\ 


N. 




^s 


\ 










\ 


% 
\ 

% 


\ 




V 




1 










X 




» 


X 


^ 
















% 


s 




















S 


s> 



e 



te 






9 



o 

0» 



4 






o x: 

.£ ^ 

o ^ 

** 6 

O 

u .t: 

S e5 

»«• 

o £ 

c 6 

C CO 

o e: 

-o 2 



§ ° 

CO •»» 

o •** 

-^ s 

15 -» 

O O 

CZ) o 

^ o 

So 

o o 



t; o o s o c 
V ^ i» lO M n 

k : 



m 

I 



00 
00 

O 
08 







C 

t 



c 

c 
> 

'S 
*5 

g 

e 
c 

c 



X 

'Z 
a 

c 



ti: .ii ?»' 



r^ ; 



9U4V4i) 



c3 
S . 



c 

5 s 



00 

J3 



C 









E. V. McCollum and M. Davis 























/ 


/ 














/ 




// 


'/ 


^/ 


^ 1 




/ 




^1 


/ 


/ 


/ 


/ 




/ 


■/. 


/ 


/ 


/ 


/ 




■/ 


■/^ 


/ 


/ 


y 


/ 






1/ 


/ 


/ 


/ 


/ 




".*;: 





C'hart IX. (6 fcmnlea.) Shows the RTOwlh of rats receiving 10 [icr 
cent of milk protein. The curves areall thoaeof fpniales. Note theeitra- 
ordinary growth of one individual. 

The composition of the ration wub as follows: 



426 Influence of Protein Intake on Growth 



tfV Jta- 







# 












7r 







^ 














4 


^ 














J'?' 


/- 


X 










« 




jj 




• 




















^ 




• 




r^ 




jr 














_y 






















**» • 




» 








• 






. 9 < 


• 
• 



f 






< 
• 
• 


• 


/- 








V 


^^ 




^0^ 


~^- 







Chart X. (5 males.) Shows the behavior of rats receiving 6 per cent 
of protein derived from the entire wheat kernel. The rate of growth 
during the first 42 days is about the same as with 4 per cent milk protein 
(Chart rV) or 4 per cent wheat embryo protein (Chart XI). 

The ration consisted of: 



per eeni 

Wheat 66.6 

Dextrin 31.5 

Butter-fat 5.0 

Salt mixture 6.9 



COMPOeiTION OP THE SALT MIXTURE 

gm. 

NuCl 1.40 

KsHPOi 2.631 

K citrate 0.740 

CaS042HiO . 578 

Ca lactate 7 068 



£. V. McCollum and M. Davis 



427 



















to 
















/* 




- — 


— 










4f0 




« 












.f* 


» 


-r 

9 












ko 


» 
» 


^ 


— 










£ 
















tt^' 




^^ 


^ 










ko 




-^ 












*ftf 
to 

60 


















/^ 


^ 










^ 


-^ 













Weens 



/o 



'4 



Chart XI. (6 males.) Shows the behavior of young rats receiving 
but 4 per cent of protein derived from wheat embryo. The growth during 
the first 42 days is comparable to that of rats getting milk protein at the 
same level (Chart IV) and with those getting the proteins of the entire 
wheat kernel 6 per cent (Chart X). 

The ration consisted of: 



percent 

Wheut embryo 13.3 

Dextria 77.4 

Bulter-fat 5.0 

Salt mixture 5.3 



coMPoerriON or tui: s.\lt mixtube 

gm. 

NaCl 1.067 

K citrate 0.205 

KtHPOi 3.016 

CaCU 0.386 

CaS042Ha 0.381 

CalaoUte 5.553 

Fe citrate 0. 100 



428 Influence of Protein Intake on Growth 



So 



j0 



<5 





• 










J 


^ 












* 
• 
• 


\l 










.^^ 


















-^ 










..^^ 








• 










■ — 


.^^ 








-^^ 



i« 



4FO 



J' 

^ ^ 6 <> ^ ^ ^ 

Chart XII. (6 males.)* Shows the steady decline of rats receiving 
2.53 per cent of protein (N X 6.25) from desiccated egg. Subtracting the 
nitrogen in the ether extract of egg (0.6 per cent) gives 2.45 per cent as 
the approximate protein content of the ration. 

The ration consisted of : 

DeaiccaUxl eg« 5.6 (45. 31% N X 6.25) 

Dextrin 87.4 

Butter>fat 5.0 

Fc citrate 0.1 



ON THE MUTAROTATION OF PHENYLOSAZONES OF 

PENTOSES AND HEXOSES. 

By p. a. LEVENE and F. B. LA FORGE. 
{From the Laboratories of the Rockefeller Institute for Medital Research.) 

(Received for publication, January 30, 1915.) 

The identification of a sugar can be made unmistakable when 
it is possible to isolate it in crystalline form, or^when there is an 
abundant supply of material at hand. The biological chemist is 
often confronted with conditions where a sugar is present in a 
minimal quantity and where the impurities predominate in a 
measure which makes the purification of the sugar an impossi- 
bility. In such circumstances one is frequently forced to base 
his opinion on the configuration of the sugar on the properties 
of the derivative which is always obtainable; namely, on that of 
the osazone. 

Naturally the same osazone is derived from two epimeric sugars, 
and when an osazone is identified there still remain two possible 
explanations for the structure of the sugar. The choice between 
the two can often be made if the information on the constants 
of the osazone is supplemented by information on the character 
of the optical rotation of the sugar. In the course of our work 
on naturally occurring rare sugars we were convinced of the 
importance of the identification of osazones, and we found that 
one of the properties which is perhaps the least afifected by slight 
impurities is the character of the mutarotation of the phenylosa- 
zone. The initial rotation of an osazone in Neuberg's pyridine- 
alcohol solution is subject to small variation, dependent on mini- 
mal impurities, but the direction of the mutarotation and the 
equilibrium rotation remains constant. 

We found this property of phcnylosazones serviceable in the 
work on the identification of the o-ketoxylose in a case of pento- 
suria.* Zemer and Waltucli,^ independently of us and later only 

* This Journaly xviii, p. 319, 1914. 

' E. Zerner and R. Waltuch: Monnish.f. Chcm., xxxv, p. 1025, 1914. 

429 

THJP JOURNAL OP Btq^^OOICAL CHClllSTRT, TOL. XX, NO. 3 



430 Mutarotation of Phenylosazones 

by a few weeks, called attention to the same property of phenyl- 
pentosazones. And again we made use of it in the work on 
chondrosamine. 

In view of this we considered it important to determine the 
character of the mutarotation of the phenylosazones of all the 
normal pentoses and hexoses. Also the melting points of the 
osazones were revised on this occasion. 

EXPERIMENTAL. 

Hexoses. 

AUrose osazone was prepared from crude altrose obtained by 
reducing altronic^acid lactone with sodium amalgam. It was 
recrystallized three times from hot water. 

(hilose osazone was obtained from gulose prepared by the 
cyanhydrin synthesis from Z-xylose. It was twice recrystallized 
from about 10 per cent acetic acid, dried, and then recrystallized 
from 4 parts of absolute alcohol plus 10 parts of ether. The 
product thus purified consists of long, yellow, hair-like needles 
which do not darken on exposure to air. 

Galactose osazone was twice recrystallized from a large volume 
of dilute alcohol and once from absolute alcohol. 

Glucose osazone was recrystallized from absolute alcohol. 

Corrected melting points of osazones at a rede of heating of about 4 to 6 

seconds per i**. 

OSAZONK8 CONTRACTED ' BELTED AT -C. ^^OUrOB^r> 

AT Ky. AT O. 

I 

• /-Arabinose 160 

I d-Xylosc 

Altrose 175 

r Gulose 

Galactose 199* 

» Glucose 

*Thc same sample ^^'as heatod 1.5 minutes at 190* without melting. 

Optical rotations. 

d'Xylosazone. I. Soon after preparation of solution [a]o = —0.10 

After about eighteen hours —0.36 

II. The same ten minutes after preparation of solution —0.09 

After one hour —0.21 

After twentv-fonr hours —0.43 



166 


200 


164 


167 


178 


189 


168 


180 


201 ' 


202 


208 


208 



p. A. Levene and F. B. La Forge 431 

/-.t rnhinoaazone. Soon after preparation of solution [a]^ - +0.55 

After about eighteen hours +0 .30 

d- A It rose osazone. Soon after preparation of solution [ajn == —0.40 

After twenty-four hours —0 .29 

d-Ouloae osazone. Soon after preparation of solution [alo = +0.07 

After twenty-four hours +0 .60 

d-Galactose osazone, I. Soon after preparation of solution [ajn = . . +0.73 

After eight hours +0.32 

II. Soon after preparation of solution +0.80 

After twenty-four hours +0 .34 

d-Gliicosc osazone. Soon after preparation of solution (ajo = —0.62 

After twenty-four hours. —0 .35 

All determinations were made in a 0.5 dm. tube with D-light, O.l gm. of 

substance in 5 cc. of pyridine-alcohol mixture being used. 



ON CHONDROITIN SULPHURIC ACID. 

FOURTH PAPER. 

By p. a. LEVENE and F. B. LA FORGE. 

(Frofti the Laboratories of the Rockefeller InstiiiUe for Medical Research.) 

(Received for publication, January 26, 1915.) 

In the third paper^ on this subject observations were reported 
which led to the conclusion that the nitrogenous component of 
chondroitin sulphuric acid was a new hexosamine. The data in 
possession at that time were insuflBcient to permit a definite 
conclusion regarding the details of the structure of the new sugar, 
which was named chondrosamine. Additional information was 
gained in the course of the present season, and today the structure 
of the sugar seems to be explained, except on one point; namely, 
on the relative position of the amino-group attached to the a- 
carbon atom. 

Chondrosamine was oxidized with bromine to tetroxyamino- 
caproic acid, which differed from glucosaminic acid in its solubility 
and in its optical rotation. On reduction with hydriodic acid 
in the presence of phosphorus, according to Fischer and Tiemann^ 
a monohydroxyaminocaproic acid was obtained which seemed to 
\^ identical with the substance obtained under the same conditions 
from glucosaminic acid. Since glucosamine is known to contain 
a normal carbon chain, it became evident that chondrosamine 
contained a carbon chain of the same character. 

The further data on the structure of the sugar were gained 
through the study of the products of oxidation of chondrosamine 
to its dicarboxylic derivative. In the last paper it was stated 
that on oxidation of desaminochondrosamine with nitric acid a 
dicarboxylic acid was obtained in the form of its calcium salt. 
The salt differed from the corresponding salt of isosaccharic acid. 

' P. A. Lcvene and F. B. La Forge: this Journalf xviii, p. 123, 1914. 
^E. Fischer and F. Tiemann: Ber. d. deutsch. chetn. Gesellsch., xxvii, 
p. 138, 1894. 

433 



/ 



434 



Chondroitin Sulphuric Acid 



It crystallized with water of crystallization, two molecules of 
which were lost on heating at 108° at diminished pressure over 
phosphorus pentoxide. On further drying for eighteen hours 
at 138°C. under diminished pressure no further loss of weight 
could be obtained. The substance then analyzed for a dicar- 
boxylic acid of a tetroxyadipic acid. However, the properties 
of the substance differed from those of any one of the dicarboxylie 
acids obtained from any one of the known hexoses. The acid 
was named chondrosic acid. It was soon discovered that in a 
free state the acid analyzed for an anhydrotetroxyadipic acid. 
None of the carboxyl groups took part in the anhydride formation. 
Hence it became evident that chondrosic acid was isomeric with 
isosaccharic acid. 

Chondrosic acid treated by the methods of Tiemann similarly 
to isosaccharic acid gave rise to a- ai-furandicarbonic (dehydro- 
mucic acid) and to furancarbonic acid (pyromucic acid). Thus 
it became known that in chondrosic acid similarly to isosaccharic 
acid an oxygen bridge was established between the a- and ai- 
carbons. 

Theoretically the niunber of possible acids of this structure is 
as great as that of the normal dicarboxylie acids of the theoretic- 
ally possible hexoses; namely, six in the dnseries and as many in 
the l' or in the d/-series. 



COOH 
HlOH 
OHH 
HOH 

h|oh 

COOH 



II 

COOH 
OHlH 
OHH 
HOH 

h|oh 

COOH 



III 

COOH 
OHlH 
OHH 

HrOH 

OHH 
COOH 



IV 

COOH 
HJOH 
OHH 

HOH 
OHH 
COOH 



COOH 
HbH 
OHH 
OHH 

h|oh 

COOH 



VI 


VII 


VIII 


COOH 


COOH 


COOH 


OHH 


OHH 


HlOH 


OHH 


HOH 


HOH 


OHH 


hIoh 
h|oh 


hIoh 
h|oh 


HOH 


COOH 


COOH 


COOH 



It is obvious that I and III, and VI and VII are identical. 




nil 



O 



OH 
H 



COOH 



COOH 



COOH 

H 

H 

OH 

H 
COOH 



IV, 

COOH 
hI-tQ 
OHP 

COOH 



H 



— O 



OHH 

OHlH 

H 



V, VIi Vll, VIII, 

COOH COOH COOH COOH 
H y—]H Hi— O 
II OH HOH 
HOH HOH 
H_\) hL_ 

COOH COOH COOH COOH 




p. A. Levene and F. B. La Forge 435 

Also here Ii and IIIi, and VIi and VIIi are identical. 

If all these anhydrodicarboxylie acids were known the identifi- 
cation of chondrosic acid should have offered no diflSculty. Un- 
fortunately isosaccharic acid was the only known member of this 
group of acids. 

Only two of the six acids or of their anhydro derivatives (mucic 
and allomucic acids) are optically inactive and can be distifi- 
guished by this property from the other acids. It was possible 
to obtain from chondrosamine an inactive anhydrodicarboxylie 
acid. 

Furthermore, the properties of the osazone* derived from chon- 
drosamine (the melting point and character of its optical rotation) 
seemed identical with those of the osazone obtained from allose 
or altrose. On the basis of this there seems only one conclusion 
possible; namely, that chondrosamine has the structure of {-a- 
allosaminie. (The oisazone of d-allose is levorotatory, that of 
chondrosamine dextrarotatory.) In reality also the structure of 
/-a-altrosamine may be ascribed to chondrosamine. The rea- 
sons are the following: Chondrosamine gives rise to the inactive 
anhydrodicarboxylie acid only under very definite conditions; 
namely, on treatment of the monocarboxylic amino-acid formed 
when the amino-sugar is oxidized with bromine, first with nitrous 
and subsequently with nitric acid. However, if chondrosamine 
is directly treated with nitrous and then oxidized with nitric 
acid, the product is an active anhydrodicarboxylie acid. 

Evidently in the process of the formation of one of the two 
anhydro acids the original structure of the sugar suffered a re- 
arrangement. In all probability this rearrangement is brought 
about by the action of nitrous acid, since there can be derived 
also two monocarboxylic acids of the desamino chondrosamine. 
On the basis of this it is permissible to assume that the rearrange- 
ment affected the groups attached to the a-carbon atom, and 
hence that the two anhydrodicarboxylie acids may be regarded, 
one as corresponding to altrose and the other to allose. 

For the present there exists no experimental evidence for a 
definite conclusion regarding the arrangement of the groups on 
the a-carbon atom in the original sugar.* 

• Levene and La Forge: loc. cit. 

* See J. C. Irvine and A. Hynd: Jour. Chem, Soc., ci, p. 1128, 1912; cv, 
p. 698, 1914. 



436 Chondroitin Sulphuric Acid 

An attempt was also made to prepare chondrosaminic acid 
synthetically by the addition of hydrocyanic acid to ribosimine, 
but the acid obtained in this manner proved to be different from 
chondrosaminic acid. Unfortunately the yield was too small to 
allow a further oxidation into the anhydrodicarboxylic acid. 
The poor yield of the substance was due principally to our un- 
familiarity with the action of prussic acid on ribosimine. The 
reaction took place with unexpected violence and the products 
warmed up to a temperature which was detrimental to the 
compound. 

In conclusion it may be mentioned that our observations throw 
additional Ught on the relationship of chitonic and chitaric acids. 
Fischer and Tiemann have shown that glucosamine yields either 
one or the other of the two monocarboxylic acids, depending on 
the procedure followed in their preparation, and they have 
explained the differences of the two acids on the basis of differences 
in configuration. It was found in the course of this work that 
also the dicarboxylic acid corresponding to chitaric acid differs 
from isosaccharic. From the analogy with chondrosamine one 
may be inclined to regard chitaric^ and chitonic acids as epimers, 
one having the configuration of glucose and the other of mannose. 

Work on the preparation of other amino-sugars and their oxi- 
dation products is now in progress. There are already prepared 
by us the derivatives of xylose and ly^tose. 

EXPERIMENTAL. 

Chondrosaminic add. 

Sixty grams of chondrosamine hydrochloric acid salt were 
dissolved in 600 cc. of water and allowed to stand for one week 
at room temperature with an excess of bromine. The oxidation 
was then continued for two weeks longer at 35*^ to 40° with 
occasional shaking, bromine being added from time to time so 
that an excess always remained on the bottom of the vessel. 
The reaction product was then concentrated to a thin syrup in 
vacuum to remove the excess of bromine and most of the hydro- 
brominic acid. It was then diluted to about 150 cc. with alcohol, 

*£. Fischer: Ber. d. deutach. chem, Geaellsch,, xxiv, p. 2130, 1891. 



p. A. Levene and F. B. La Forge 437 

which caused the separation of the unchanged amino-hcxoses as 
hydrobromic acid salt. After standing for a few hours in the 
refrigerator the alcoholic solution was poured off, the residue 
extracted with a little alcohol, and filtered. The combined 
alcoholic solutions were then concentrated in vacuum to a small 
volume, taken up in 400 cc. of water, and the halogen removed 
first with lead carbonate and finally by warming on the water 
bath with silver carbonate. The silver and lead were removed 
from the filtrate with hydrogen sulphide and the solution was 
warmed for an hour with animal charcoal and filtered. The 
filtrate, which was of a pinkish color, was then concentrated to 
a small volume in vacuum when crystallization took place spon- 
taneously. The crystals were washed out of the flask with a 
small amount of 50 per cent methyl alcohol, filtered with suction, 
and washed on the funnel with 50 per cent methyl alcohol until 
nearly colorless. The yield of the crude product, dried in vacuum, 
was 15 grams; while 4 grams more were obtained by concentrating 
the mother liquor. About 10 grams of the hexosamine hydro- 
bromic acid salt were recovered from the oxidation product. 
The substance crystallizes from 10 per cent aqueous solution or 
from 50 per cent alcohol in short prismatic needles grouped to- 
gether in rosettes. It has no melting point but darkens slowly 
above 190**. 

0.1396 gm. of substance gave 17.3 cc. amino N, at 22^ 760 nun. 
0.0992 gm. of substance gave 0.1326 gm. CO2 and 0.0568 gm. HsO. 

Calculated tor Found: 

CHijOiN: 

C 36.92 36 .61 

H 6.66 6.44 

N 7.18 7.02 

0.1498 gm. of substance in 2 cc. of 2.5 per cent HCl rotated in a 1 dm. tube 

with D-light at 28.5®, shortly after preparation of the solution -1.21®; 

at 25®, after 48 hours -2.19®. 

[a]"* = -16.15®, -29.23® 

Redixtion of chondrosaminic acid with hydriodic add. 

Two grams of substance were heated with 18 cc. of hydriodic 
acid, specific gravity 1.96, and 5 grams of phosphonium iodide 
for four and a half hours in a sealed tube at 100°. The contents 
of the tube which remained colorless were diluted to 500 cc. with 



438 Chondroitin Sulphuric Acid 

water and the hydriodie and phosphoric acids removed with 
lead carbonate. The filtrate which contained halogen was heated 
on the water bath with silver carbonate, filtered, treated with 
hydrogen sulphide, filtered, and the filtrate from silver sul- 
phide concentrated in vacuum to dryness. The residue was 
dissolved in 10 cc. of water and concentrated on the water bath 
to a thick syrup which solidified on cooling to a semisolid cake. 
This was extracted with 5 cc. of methyl alcohol, filtered, and dried. 
The yield was 0.8 gram. For analysis it was dissolved in 2 cc. 
of water, diluted with 10 cc. of methyl alcohol, filtered from a 
light flocculent precipitate and concentrated in vacuum to a 
small volume. The crystals thus formed were filtered with suction 
and washed with methyl alcohol. The substance contracted at 
217'' and melted with gas evolution at 222° to 223** (uncorrected). 

0.1199 gm. of substance gave 20.5 cc. amino N, at 26**, 763 mm. 
0.1026 gm. of substance gave 0.1828 gm. CO2 and 0.0792 gm. H2O. 

Calculated for Found : 

CcHi^N: 

C 48.98 48.98 

H .• 8.84 8.65 

N 9.52 9.48 

Chondrosic add, 

6.8 grams of the calcium salt described in a previous pap(»r* 
were introduced in portions into 250 cc. of boiling water con- 
taining 3.2 grams of oxalic acid. After one-half hour's boiling 
the solution was treated with a Httle animal charcoal and filtered, 
the filtrate concentrated to a syrup which was dissolved in about 
30 cc. of acetone, filtered, and again concentrated to a syrup on 
the water bath. On standing over night large colorless prisms 
had separated out. Upon completion of crystallization the 
adhering syrup was dissolved from the crystals with a few cc. of 
a solution of equal parts of acetone and ether. The crystals 
were then filtered off with suction and washed with a small amount 

' Jjevenc and La Forge: loc. cii.f p. 128. Instead of allowing the 
deaminized solution of the hcxosamine to stand over night at 42^ with 
nitric acid, we mixed the solution with nitric acid at 0^ and allowed it to 
stand at room temperature. It was then treated as before and we were 
able to obtain constantly a yield of from 50 to 60 per cent calcium salt. 



p. A. Levene and F. B. La Forge 439 

of the acetone ether solution. The crystals consisted of large 
parallelopipeds which melted at 179° to 181** (uncorrected) and 
at this temperature slowly decomposed with gas evolution. The 
pure substance is rather diflScultly soluble in acetone (about I 
to 40 at the boiling point of the solvent). For analysis it was 
dissolved in acetone, which was allowed to evaporate partially. 
The melting point of the recrystallized product did not differ 
from that of the first. The substance is easily soluble in water 
and alcohol, but nearly insoluble in ether. The yield amounted 
to 2.2 grams. 

0.1094 gm. of substance gave 0.1480 gm. CO2 and 0.0404 gm. HtO. 

Calculated for Found: 

CsHiOr: 

C 37.50 37.11 

H 4.20 4.15 

0.1500 gm. of substance required 15.9 cc. -nr NaOH (calculated: 15.6 cc). 
0.2004 gm. of substance in 3 cc. HtO, weight of solution 3.1835 gm., 
specific gravity 1.0262, rotaled in a 1 dm. tube with D-light at 28^ -1.07% 
and was unchanged after 5 hours. 

[a]" - -16.56« 

Epichondrosic add. 

Fifteen grams of chondrosaminic acid dissolved in 90 cc. of 3.3 
per cent hydrochloric acid were deaminized with 15 grams of 
silver nitrite, first at 0° for two hours and then for twenty hours 
at room ^temperature, 4 grams more of silver nitrite being added 
towards the end of the experiment, together with a few cc. of 10 
per cent hydrochloric acid. The excess of silver was removed 
from the filtrate with hydrogen sulphide and the solution con- 
centrated in vacuum to about 35 cc. An equal volume of con- 
centrated nitric acid was then added in the cold and the mixture 
evaporated to a thick syrup on a large watch glass in two portions 
on the water bath. This was taken up in a few cc. of water and 
again evaporated to a syrup. Crystallization occurred spontane- 
ously during the second evaporation. After standing for a short 
time the crystals were freed from the syrupy by-products by 
washing with alcohol and ether mixture. When filtered and dried 
the yield of the first product amounted to 6.8 grams. It was 
recrjTstallized by dissolving in a large amount of acetone and 



440 Chondroitin Sulphuric Acid 

evaporating the solution to a small volume. Melting point: 
201° to 202° (uncorrected). 

0.1018 gm. of substance gave 0.1402 gm. CO2 and 0.0402 gm. H3O. 

Calculated for ' Found : 
C^HfOr: 

C 37.50 37.56 

H 4.20 4.42 

0.1137 gm. of substance required 11.7 cc. iV NaOH (calculated: 11.8 cc). 

0.2000 gm. of substance dissolved in 2 cc. of water showed no appreciable 
rotation in a 1 dm. tube with D-light, measured under conditions where a 
rotation of 0.02^ could not have oscaped detection. 

Dehydromudc add from chondrosic acid. 

One gram of chondrosic acid in 1 cc. of concentrated hydro- 
chloric acid plus 1 cc. of concentrated hydrobromic acid heated 
in a sealed tube at 150° for eight hours, according to Fischer*'' 
gave 0.2 gram of dehydromucic acid. 

0.1002 gm. of substance gave 0.1684 gm. COs and 0.0276 gm. HsO. 

Calculated for Found: 

CHiO.: 

C 46.16 45.84 

H 2.57 3.08 

Pyromucic acid from chondrosic acid. 

One gram of substance was heated at about 200° in a test- 
tube in an atmosphere of carbon dioxide for one-half hour. The 
substance sublimed and separated out on the cooler pai:ts of the 
tube from which it was afterwards removed, extracted with 
ether, and recrystallized by dissolving in a large amount of ether 
and allowing the solution to evaporate. It melted at 135°. 

0.1000 gm. of substance gave 0.1964 gm. CO2 and 0.0370 gm. HtO. 

Calculated for Found: 

CiHiOi: 

C 53.57 53.56 

H 3.57 4.14 

RibosimineJ 

Twenty-five grams of crystalline d-ribose were dissolved in 
about 25 cc. of saturated dry methyl alcohoUc ammonia. Crystal- 

" C. A. Lobry de Bruyn and F. H. Van Leent: Rec. d. trav, chim. d 
PayS'Bas, xiv, p. 134, 1895. 



p. A. Levene and F. B. La Forge 441 

lization began after two days and was complete in about three 
days. The hard crystalline crusts were broken up, filtered, and 
washed with dry methyl alcohol. The yield of dried substance 
was 22.5 grams. It melted at 137** to 138** (uncorrected) with 
decomposition. 

0.1990 gm. of substance gave 13.4 cc.-ri HCl (Kjeldahl). 

Calculated for Found: 

C»HuN04: 

N 9.40 9.42 

Hexosaminic acid from ribose. 

Twenty grams of ribosimine were covered with 10 cc. of water® 
and 6 cc. of 80 per cent hydrocyanic acid added. The reaction 
proceeded almost violently, and cooling in a freezing mixture 
had to be resorted to. After about fifteen minutes, the product, 
a thick brown syrup, was dissolved in about 100 cc. of cold con- 
centrated hydrochloric acid and allowed to stand for twenty- 
four hours at room temperature. The solution was then con- 
centrated in vacuum to a small volume, diluted, and the ammonia 
completely removed by distilling in vacuum with an excess of 
barium hydrate. The barium was removed with a slight excess 
of sulphuric acid. I^ead carbonate was then added to neutral reac- 
tion and the precipitates of barium sulphate and lead chloride were 
filtered off together. The filtrate was warmed with silver carbon- 
ate to remove the rest of the halogen, filtered, and the lead and 
silver were removed from the filtrate with hydrogen sulphide. 
The solution was then concentrated to a small volume. Upon 
addition of alcohol a thick brown syrup was precipitated. The 
supernatant liquid was poured off, and as the syrup failed to 
crj'stallize after two weeks' standing, this was extracted several 
times with about 15 per cent methyl alcoholic hydrochloric acid, 
and thus a separation from the thick amorphous by-product 
effected. The alcoholic extract, which was nearly colorless, Wiis 
concentrated in vacuum to a small volume which was taken up 
in water and tlie hydrochloric acid removed by warming on the 

* As the subsequent experiment shows, this concentration was too 
great. The* experiment should be made in greater dilution, but lack of 
sufficient ribose prevented a repetition of the experiment for the time 
being. 



442 Chondroitin Sulphuric Acid 

water bath with silver carbonate. The silver was removed from 
the filtrate while still hot and the solution concentrated to a syrup 
which was treated with a large volume of methyl alcohol. The 
resulting nearly colorless syrup finally sohdified upon standing. 
The semisolid cake thus obtained was extracted with a few cc. of 
methyl alcohol and filtered. It was then dissolved in 4 to 5 parts 
of hot water with the addition of a little animal charcoal and 
filtered^ To the filtrate an equal volume of methyl alcohol 
was added. On standing the substance crystalUzed in short 
prismatic needles grouped together in rosettes. The yield was 
only about 0.5 gram of the pure substance, which melted with 
decomposition at 198° (uncorrected). 

0.0990 gm. of substance gave 0.1344 gm. CO* and 0.0582 gm. HjO. 
0.0261 gm. of substance gave 3.45 cc. amino N, at 22®, 756 mm. (micro 
method of Van Slyke). 

Calculated for Found: 

C«Hi«0«N: 

C 36 .92 37 .02 

H 6.66 6.58 

N 7.18 7.4^ 

0.1500 gm. of substance in 2 cc. 2.5 per cent UCl, weight of solution 
2.1663 gm., rotated in a 1 dm. tube with D-light at 28**, after 20 minutes 
-1.67**; after 24 hours -0.71**; after 48 hours -0.71**. 

[a]"=-22.26«,-9.4* 

Epi'isosaccharic acid. 

Ten grams of glucosaminic acid were deaminized with silver 
nitrite according to Fischer and Tiemann," the resulting solution 
concentrated to about 25 cc, and an equal volume of nitric acid 
added in the cold. Oxidation was carried out under the same 
conditions as already described for chondrosaminic acid. The 
syrupy product was dissolved in 10 cc. of water, neutralized with 
strong potassium hydrate solution, and an equal volume of 
glacial acetic acid added. Upon dilution to 50 cc. with alcohol 
the acid potassium salt of the dibasic acid soon began to crystal- 
lize, and after standing one-half hour in the refrigerator the yield 
of the substance washed and dried amounted to about 4 grams. 
For analysis it was recr\'stallized from about 2 parts of hot water. 

* Fischer and Tiemann: loc. cit., p. 145. 



p. A. Levene and F. B. La Forge 443 

The product contains 0.5 of a molecule of crystal water, which 
can be removed in vacuum at 115°. 

0.1968 gm. of substance gave 0.0058 gm. UtO. 

0.1811 gm. of dried substance gave 0.0511 gm. KsCO|. 

Calculated for Found: 

CtHffOtK + 0.5 HtO: 

H,0 3.7 3.10 

Calculated for 
C«HfOiK: 

K 15 .73 15 .94 

Lead salt of epi-isasaccharic acid. 

2.5 grams of the acid potassium salt were dissolved in about 
100 cc, of water and an excess of neutral acetate solution was 
slowly added in the cold. Crystallization b^an after a few 
minutes and was complete after one-half hour. The lead salt 
crystallizes under these conditions in large, nearly square plates 
which are very diflScultly soluble in water. The yield was practi- 
cally quantitative. The salt contained 2 molecules of crystal 
water, which could be removed by heating in vacuum at 150°. 
The dried preparation analyzed for a neutral lead salt of an an- 
hydrodicarboxylic hexonic acid. 

0.2418 gm. of substance gave 0.0204 gm. H2O. 

C-alculated tor Calculated for Found: 
C«HtOiPb+2HiO: CfH*OTPb+2HtO: 

H,0 7.99 8.32 8.40 

0.2214 gm. of dried substance gave 0.1690 gm. PbSOi. 

Calculated for Calculated for Found: 
CHtOtPb: C«H«07Pb: 

Pb 49.88 52.18 52.10 

7.8 grams of the lead salt were suspended in 150 cc. of water 
and slightly less than the calculated amount of sulphuric acid 
was added. The suspension was warmed on the water bath for 
one hour and then diluted with an equal volume of alcohol, fil- 
tered, and the filtrate concentrated in vacuum to a thick syrup. 
This was dissolved in acetone, filtered, and concentrated on the 
water bath. The resulting syrup crystallized in the desiccator 
to a semisolid cake. The crude product was extracted with a 
small amount of a mixture of 1 part amyl alcohol and 2 parts of 
ether, filtered, and washed with the same solvent mixture and 



444 ' Chondroitin Sulphuric Acid 

finally with dr>'^ ether. It may be recrystallizcd by dissolving in 
a small amount of acetone and allowing the solution to evaporate 
nearly to dryness in the air. It crj'^stallizes in aggregates of large 
plates, the edges of which are usually rounded off by the solvent 
action of the ether used for washing. The acid is extremely 
soluble in the usual reagents, with the exception of cold amyl 
alcohol and ether. Yield, about 2 grams of pure substance. The 
product contains 1 molecule of crystal water which can be removed 
by heating in vacuo at 78°. The dry substance melts at 160° 
(uncorrected). 

0.1278 gin. of Bubstance gave 0.0107 gm. HjO. 

Calculated for 
CeHiOr+HjO: Found: 

H,0 8.57 8.38 

0.1159 gin. of substance gave 0.1598 gm. COj and 0.0467 gm. H^O. 

Calculated for 

C^HsOt: Found : 

C '. . 37.50 37.57 

H 4.28 4.47 

0.0898 gm. crystal water containing substance required 8.6 cc. {o 
NaOH (calculated: 8.4 cc). 

0.1637 gm. of dry substance in 2 cc. of HjO, specific gravity 1.03^i, 
rotated in a 1 dm. tube at 28^ vnth D-light +3.12**. 

Mt = +39.70 



THE BEHAVIOR OF BACTEMA TOWARDS PURIFIED 
ANIMAL AND VEGETABLE PROTEINS. 

By JOEL A. SPERRY and LEO F. RETTGER. 

{From the Sheffield Laboratory of Bacteriology and Hygiene, Yale University^ 

New Haven.) 

(Received for publication, February 2, 1915.) 

Bacterial decomposition of proteins has been the subject of 
numerous and extensive investigations.. Among those who have 
made important contributions in this field are Hauser, Salkowski, 
Hirschler, Nencki, Wintemitz, Wemick, Salus, Stockley, Bien- 
stock, and Tissier. Little attention has been given, however, 
to the decomposition of pure proteins in media from which all 
other nitrogen-containing substances were excluded. 

In the painstaking investigations of Bainbridge^ sufficient 
evidence was acquired to indicate that certain aerobic and facul- 
tative anaerobic bacteria are in themselves unable to initiate or 
bring about the decomposition of purified unaltered animal pro- 
teins. In other words, it was shown that media which contain 
native proteins as the only nitrogenous ingredients fail to furnish 
the necessary conditions of bacterial development, although all 
the other requirements of a culture medium are satisfied. The 
investigations of Bainbridge were limited to a relatively small 
number of bacteria, not including the so called '* putrefaction 
organisms," and the proteins egg albumin and serum albumin 
alone were employed. 

The present investigation is in part a repetition of the work of 
Bainbridge. In addition, however, considerable attention was 
devoted to putrefactive anaerobes of the putrificus and mulignnnt 
edema bacillus type. Furthermore, the deportment of bacteria 
towards the vegetable protein edestin was studied; and, finally, 
a number of experiments were conducted with a \4ew of determin- 
ing the role of bacterial enzymes in the so called *' proteolytic 
activities" of bacteria in nature. • 

' F. A. Bainbridge: Jour, Hyg.y xi, p. 841. 1911. 

445 

TBKiOURITALOr BIOLOaiCALCHKMnTRT. VOL. XX. NO. 3 



446 Behavior of Bacteria Towards Proteins 

Crystallized egg albumin was prepared by the method of Hopkins and 
Pinkus.* After two or three recrystallizations and at least ten precipi- 
tations with powdered sodium sulphate, the albumin solution was subjected 
to dialysis until the dialysate ceased to give a reaction for ammonia with 
Nesslcr's reagent. This required from five to seven days. 

The per cent of nitrogen present in the dialyzed albumin solution was 
determined by the Kjeldahl method, and from this the amount of protein 
was estimated by multiplying the nitrogen figure by 6i25. The protein 
solution was then diluted with ammonia-free water until it contained 
0.2o to 0.5 per cent protein. The following inorganic salts were added: 
sodium chloride 0.5 per cent, sodium sulphate 0.1 to 0.2 per cent, calcium 
chloride 0.1 per cent, and potassium phosphate 0.1 per cent. The final 
reaction was adjusted so that the solution was faintly acid to litmus. 

Sterilization of the pure-protein medium was accomplished by Berke- 
feld filtration, which appears to be the only method that will insure the 
albumin against any action that may change its chemical composition. 
The Berkefeld filter and the receiving bottle were sterilized by heating in 
an autoclave for one hour under an extra pressure of ten to twelve pounds. 
The sterile filtrate was introduced in accurately measured quantities into 
sterile test-tubes and flasks. Ten cc. of the albumin solutions were trans- 
ferred to the test-tubes, while the flasks received 25 cc. Sterility tests 
were always made by streaking slant agar tubes with several loopfulls of 
the solutions and incubating the agar for 24 hours at 37%^. and for an ad- 
ditional period of 24 to 36 hours at 24°C. 

The sterile solutions of albumin and inorganic salts were inoculated 
with 24 hour slant agar cultures of the various organisms. An extremely 
small amount of the growth was removed from the surface of the agar 
with a platinum needle, great care being exercised to prevent the transfer 
of any of the agar. After thoroughly shaking the test-tubes or flasks 
definite dilutions were made as follows: One cc. of the bacterial suspension 
was mixed with 100 cc. of sterile water in dilution bottles. This consti- 
tuted dilution A (1': 100). One cc. of dilution A was further diluted by 
mixing it with 100 cc. of water (dilution B, 1: 10,000). Agar plates were 
poured with 0.5 cc. of the last dilution. Duplicate series of plates were 
always prepared, and the results usually averaged. The plates were 
incubated at 24** or ST^C, according to the organism under observation. 
Counting of the colonies was done on the Wolfhugel apparatus, with the 
help of a magnifier. 

The technique employed in the preparation of the solutions of serum 
albumin was essentially the same as that already described for egg albumin. 
Owing to the difficulty of obtaining fresh horse serum or serum that was 
free from chemical preservatives, the number of tests made with serum 
albumin was comparatively small. Furthermore, the crystallization of 
serum albumin was attended with considerably more difficulty than that 
of the egg albumfc. 

* F. G. Hopkins and S. N. Pinkus: Jour. Physiol., xxiii, pp. 130-36, 
1898-99. 



J. A. Sperry and L. F. Rettger 



447 



Etlcstin WI18 prepared from hemp-seed, as described by Osborne.' After 
repeated dialysis the precipitate in the dialyzer was collected on filter 
paper, washed, and introduced into a flask containinf< ammonia-free water. 
The amount of protein present in the suspension after thorough shaking 
was determined by the Kjeldahl method. 

A definite portion of the edestin suspension was sufficiently diluted with 
ammonia-free water to reduce the edestin content to 0.5 to 1.0 per cent. 
Sodium chloride (0.5 per cent) and sodium sulphate (0.2 per cent) were 
added, and the medium was rendered faintly alkaline to litmus paper 
with a very weak solution of sodium hydroxide. Complete solution was 
brought about by the addition of the alkali, yet the alkalinity was so slight 
as to offer little, if any, resistance to bacterial development. Sterilization 
was effected by Berkefeld filtration, as in the preparation of egg and serum 
albumin solutions. Sterile calcium chloride and potassium phosphate were 
added to the filtered solution in requisite amounts (0.1 per cent). These 
two agents were added after the filtration so as to avoid any interference 
of them with the filtration, and to prevent the loss of calcium phosphate 
from the medium. 

The final solution containing the purified edestin was transferred to 
tcjst-tubes and was tested for sterility before being inoculated. Coagu- 
lation tests were not made with the edestin solutions, but the same quan- 
titative bacterial determinations on agar plates were made as in the study 
of the deportment of bacteria in solutions containing purified animal 
albumin. 

The behavior of aerobes and facultative anuerobes in solxUiona 
amtaining purified egg albuminy serum albumin^ and the vegetable 
protein edestin as the only available sources of nitrogen. 

The following organisms were employed: 



Ji. subtilis 

H. anthracia 

B. pyocyaneus 

H. prodigiosus 

li. prate us vulgaris (two strains) 

B. prole us mirabilis 



Bacillus Z (a rapid gelatin liquefier 
resembling B. proteus vulgaris) 
B. coll commxinis 
B. typhi (two strains) 
B, pullorum (two strains) 
S. pyogenes aureus. 



The figures in the accompilnying tables (I, III, ami IV) are 
the averages obtained from duplicate sets of plates which were 
prepared from the inoculated media at definite intervals follow- 
ing the inoculation. The small numbers of colonies of B, subtilis 
and B, anthracis that were obtained on the agar plates may be 
attributed largely to the fact that both of these organisms form 



* T. B. Osborne: Abd^rhaldens Handb. d. hiochem, Arheit^nxeihodenj ii, 
p. 289, 1910. 



448 Behavior of Bacteria Towards Proteins 



TABLE I. 

Showing the behavior of bacteria in a mediuvi containing pure egg alhujuin. 
Composition of medium: egg albumin 0,2 to 0.8 per cent, sodium chloride 
0.5 per cent, sodium sulphate O.B per cent, calcium chloride 0.1 per ant, 
and acid potassium phosphate OA per cent. 

NO. OP COLONXB8 DBVELOPINO ON AQAB PLATE8. 
AVEBAQSfl. (final DILT7TION OP OBIGINAL 

SUSPENSION 1:10,000) 
Pouring of platM after inoculation of albumin solut ion 



ORUANI8MS 



dStely' ^* ^"- ** ^"* ^^ *^- [*"* dya.,1-2 wks., 3 wks. 



B.subtilis 176 

B, subtilis 

B, subtilis 

B, subtilis 



B, anthracis. 
B. anthracis. 
B. anthracis. 



— 76 — 



B, pyocyaneus 

B. pyocyaneus 

B, pyocyaneus 125 

B. pyocyaneus 58 

B. prodigiosus 126 

B. proteus vulgaris (1) 1 

B. proteus vulgaris (1) 16 

B. proteus vulgaris (2) 83 

B, proteus mirabilis 1 



1 


1 





1 


1 


1 





2 


1 


2 





4 


398 


225 


5 


368 



— I 16 ' 
9 — 1 

— 2 







1 — 



3 , — 



696 — 

8 — 

492 — 

■ 

4 I 

2! - 



785 : — 
488 : — 

— 1209 

— 53 

166! — 



4 
3 

1 

976 

68 



567 







Bacillus Z ■ 290 

Bacillus Z ! 15 



Bacillus Z i 328 



B. coli communis count-| 457 

less 
B. coli communis ' 494 



1215 — 
1467 — 



82^1 — 
501 



2809 , 
too 
many! 
to I 
count 



— 103 







1062 — I — — 



B. typhi (1) 23 ; 

B. typhi (2) ' 540 , 

B. typhi (2) 166 , 

B. typhi (2) • 330 : 



120 

113 

o 

20 



177 
117 



443 

181 

68 
42 

mm 

44 



10 



1518 



I — — — 







J, A. Sperry and L. F. Rettger 



449 



ORQANISaCS 



TABLE J— Concluded. 

NO. OF COLONIES DEVELOPINO ON AGAR PLATEH. 
AVERAGES. (FINAL DILUTION OF ORIGINAL 

SUSPENSION 1:10.000) 
Pouriiic of plates after inoculation of albumin solution 



Imme- 
diately 



24hrs. 



48 hrs. I 72 hrs. ;4-6 dys. 1-2 wkfl. 3 wks. 



Ji. pullorum (1) 
B. pullorum (2) 



■ -| i 

89 211 — — 



9 — 



S. pyogenes aureus 27 65 

S, pyogenes aureus 165 58 

S. pyogenes aureus 154 135 

S. pyogenes aureus 84 23 



— 11 

— 3 

12 — 



— 9 — 

— — 

— — 

— ' 1 — 
2 — — 

— — 



clumps in culture media and that a thorough distribution of them 
in the agar at the time of pouring was impossible. In a few in- 
stances the first set of plates showed no colonies, while, with one 
exception, in subsequent sets some colonies appeared. Such 
irregularities must be expected with these organisms. 

Since the results are clearly set forth in the tables, it is un- 
necessary to comment on them at great length. It will be seen 
at a glance that little bacterial development took place in any 
of the pure-protein media. Control experiments in which bouil- 
lon was inoculated at the same time and in the same manner as 
the albumin and edestin solutions were conducted. Further- 
more, cultural tests were made with solutions containing a small 
amount (0.1 to 0.2 per cent) of Witte's peptone, along with the 
regular ingredients of the pure-protein media. In both cases 
bacterial development was rapid and the organisms multiplied 
to such an extent as to be apparent to the unaided eye, by the 
turbidity produced. Odor and microscopic tests further sub- 
stantiated this fact. 

While in some of the experiments an appreciable increase is 
registered in the number of colonies after the first plate pouring, 
the difference is usually offset by a decrease in other tests with 
the same bacteria. This is particularly true of B, pyocyaneiui 
and Bacillus Z. The poor showing that is made by B. proteus 
vulgaris^ in the test medium is perhaps most surprising. This 
organism, which has frequently been associated with rapid pro- 
teolysis and even putrefaction, is very quickly reduced in numbers, 



450 Behavior of Bacteria Towards Proteins 



TABLE IT. 



Control experiments showing the behavior oj bacteria in a nitrogen-free medium 
containing the following inorganic salts: sodium chloride 0.5 per centj 
sodium sulphate 0.2 per centy calcium chloride 0.1 per centj and acid 
potassium phosphate 0.1 per cent. 



so. OF COIX)NIE8 DEVELOPING ON AGAR 

PLATBa. AVERAQKS. (FINAL. DILUTION OF 

BUBPENBION 1:20,000) 



OKUANI8M8 


• 

Pouring of platee after inoculation of iho 
salt medium 




Imme- 
diately 


24 hn. 

1 


48 hra. 72 hra. 
— '■ 


4-6 dys. 


1-2 wks. 


B. suhtilis 








B. suhtilis 


134 


12 — — 1 Very 


few 






! suhtilis 






1 


colonies. 


• 






Plates con- 




1 1 
1 


taminated. 


B. suhtilis 


1 


1 







B. suhtilis 


110 








B. anthracis 


' 2 5 ' 1 




3 


B. anthracis 


4 ' 5 : — 







6 


B. pyocyaneus 


'j 215 589 ' — 37 — 

1 1 


71 


B. pyocyaneus 


6 411 ; — — 237 




B, pyocyaneus 


244 15 — 1 — 622 

t -_ i _ .1. 




B, proteu^ vulgaris 


i — 1 — 





B. proieus mirab^lis 


l_ \ \- 1- — l_ 
' 3 , ' — 





Bacillus Z 


41 75 1 — — 67 

397 ' 201 ' 




Bacillus Z 


134 




_ . _ '_ _ 1 ._ 


. 


- 




B. typhi 


368 ; 25 




123 


1 


S. vyoaenes aureus 


40 7 


1 

' n 





S. pyogenes aureus 


44 16 — — 




S. pyogenes aureus 


169 — , — 






and disappears within forty-eiRht hours after the inoculation. 
The inhibiting influence is even more pronounced than that of 
the control medium of inorganic salts (Table II). Staphy- 
lococcus aureus also undergoes rapid depletion in each of the 
pure-protein media. B. coli and B, typhi sufTered marked and 
progressive decreases except in the medium containing the pure- 



J. A. Sperry and L. F. Rettger 



451 



protein edestin, in which B. coli made slight but insignificant 
progress. 

The edestin medium was also somewhat less inhibitive or non- 
supporting than the egg or serum albumin for B. subtilisy B. 
anthraciSf and B. pyocyaneus. The differences were but unim- 
portant, however, and hence it may be safely stated that this 

TABLE III. 

Showing the behainor of bacteria in solutions of serum albumin. Compost' 
tion of mediujn: serum albumin 0,2 per cent, sodium chloride 0.6 per cent, 
sodium sulphate O.B per cent, calciujn chloride 0.1 per cent, and potassium 
phosphate 0.1 per cent. 

NO. OF COLONIES ON AOAR PLATES. AVERAGES. 
I (FINAL DILUTION OF BACTERIAL 8US- 

I PENSION 1:20,000) 



ORGANISMS 



Pouring of plates after inoculation 
of medium 



Immediately 

B. subtilis I 11 

B. subtilis 17 

_ . . _ _ _ ._ . . I ____^. 

B. anthracis , 2 

B. anthracis 2 

B. pyocyaneus 2366 

B. pyocyaneus 1801 

B. proteus vulgaris (1) 69 

B. proteus vulgaris (2) 50 

Bacillus Z i 379 

Bacillus Z 522 

B. coli communis 185 

B. coli communis 185 



24hr8. 



48hrH. 



25 


20 


23 


30 


14 


15 


19 


16 


2479 


1971 


2709 


1582 


3 


1 


3 


1 


1675 


1373 


2440 


945 


718 


492 


749 


542 



vegetable globulin, like egg and serum albumin, resists direct 
decomposition by the microorganisms employed. 

The failure to develop perceptively in these pure-protein media 
undoubtedly lies in the fact that the bacteria are not suppHed 
with nitrogen which is available for their immediate use. In 
other words, they come to a standstill in their development or 
they die from inanition. This view is supported by the control 
experiments (Table II) in which a modium consisting purely of 



452 Behavior of Bacteria Towards Proteins 

inorganic salts is used, and in which the results are practically 
the same as in the experiments involving the use of the same 
salts plus the pure protein (Tables I, III, and IV) ; and it is made 
all the more certain by the following experiments. A medium 
containing the ingredients of Uschinsky's medium, together with 



TABLE IV. 



Showing the behavior of bacteria in solutions containing edestin. Composi- 
Hon of medium: edestin 0.6 to 1,0 per cent, sodium chloride 0.6 per cent^ 
sodium sulphate O.t per cent, cdlcium chloride 0.1 per centj and potassium 
phosphate 0.1 per cent. 



ORGAKISIIB 



XO. or COLONIES DBVBLOPINQ OK AGAR PLATES. 

AVBBAOBS 



Pouring of plates after inoculation of edestin solutions 



JSJJl^j 24hrs. 48hrB. 



72hr8. M-6d>'8.1-2wks. 



3 wks. 



B. subtilis. 



2 



17 



62 I — — ! — , - 



B. anthrads. 
B. anthrads. 
B. anthrads. 



B. pyocyaneus 1678 2453 



1 





28 ' 




_— 


1 




26 


25 


37 




— 


33 


— 


11 


6 


4 , 




1 


1 : 

1 




.78 


2453 






— 


' 4346 




il7 


2903 


4391 


— 


1 


5115 





B . prodigiosus , 769 1313 1940 — i — — — 

B. prodigiosus 308 937 1593 — — i — — 



B. proteus vulgaris (1) 
B. proteus vulgaris (1), 
B. proteus vulgaris (2) 



4 

1 








Bacillus Z ' 1099 2248 

I 

B. coll communis 1042 964 

B. colt communis ' 1086 2312 







2072 

734 







— — ' — 



— — 352 — 



— 2 



S. pyogenes aureus 1051 1253 — — — 



10 



0.5 per cent pure edestin, was inoculated with the different test 
organisms and incubated for definite periods. Bacterial develop- 
ment was so rapid that in comparatively few hours the medium 
became clouded, and the bacteria were too numerous for determi- 
nation by the plate method. 



J. A. Sperry and L. F. Rettger 453 

Coagtdaiion tests for unchanged proteins. 

In order to obtain still more complete evidence as to whether 
bacteria are in themselves able to attack native proteins, experi- 
ments were conducted which involved the estimation of coagu- 
lable protein after different periods of incubation following the 
inoculation of the pure-protein medium. The method employed 
was as follows: Flasks containing 25 cc. of the test egg albumin 
solution were inoculated with B. anthracis, B. proteus vulgaris, 
and Bacillus Z. After an incubation period of from sixteen to 

TABLE v. 

Showing the resvliB o/ coagulation tests for unchanged proteins. Compost^ 
lion of protein medium: pure egg albumin 0.8 per cent, sodium chloride 
0.5 per cent, sodium sulphate O.t per cent, calcium chloride 0.1 per cent, 
and potassium phosphate 0.1 per cent. 26 cc. of medium in each flask. 

NO. OP I r»»ftA^-T«xfii 'n;^tr,„ WEIGHT OF COAGULATED PROTIIN AT END 

FLABK PERIOD INCUBATION PERIOD 



I dift. I gm. 

3 None (Control) 16 10 .0819 

4 ' None (Control) 16 | 0.0815 

5 Bacillus Z 19 0.0560 Medium contaminated. 

6 ! Bacillus Z 19 | .0679 Medium contaminated. 

7 [ Bacillus Z 17 0.0819 

8 Bacillus Z 17 0.0815 

9 B. proUus vulgaris ' 18 0.0856 

10 B. proteus vulgaris 18 ' 0.0829 

11 j B. anthracis i 23 0.0784 

12 B. anthracis 23 0.0835 



twenty-three days the contents of the flasks were made faintly 
acid with acetic acid and heated in a water bath (about 99°C.) 
for fifteen to twenty minutes. The coagulum was collected in a 
weighed Gooch crucible containing a thin asbestos mat, and 
dried to constant weight. 

The results as given in Table V show that there was no real 
loss of coagulable protein, barring two exceptions. Two dupH- 
cate flasks which were inoculated with Bacillus Z registered a 
loss of normal protein, the one about 17 per cent and the other 
31.6 per cent. Both of these flasks were badly contaminated, 
which fact undoubtedly offers an explanation of the partial 



454 Behavior of Bacteria Towards Proteins 

decomposition of the proteins. The associated bacteria were 
present in sufficient numbers to furnish available nitrogen, either 
directly or indhectly, for the test organism and the development 
of a requisite amount of enzyme to bring about proteolysis. 
Further experiments with Bacillus Z in which contamination was 
excluded showed no decomposition whatever of the egg albumin. 
These results are in perfect accord with those of Bainbridge, 
and substantiate those which were obtained earlier in our 
investigation. 

The behavior of the putrefactive anaerobes towards pure animal 

and vegetable proteins. 

The meaning of the term putrefaction and the organisms which 
cause putrefactive decomposition have been the subject of so 
much discussion in recent years that a brief reference to some of 
the opinions and theories can not appear out of place here. 

In the controversies two important views of the biology and 
chemistry of putrefaction have been strongly defended. Accord- 
ing to the older theory of Hauser and many other investigators, 
and the one still held by Tissier* and Mart^lly, MetchnikofT,^ 
and others, putrefaction is caused not only by obligate anaerobes, 
but by facultative aerobes, and even a limited number of obli- 
gate aerobes. On the other hand, it is claimed by Bienstock^ and 
Rcttger^ that real putrefaction is the work of certain anaerobes 
only. In an article published quite recently Tissier is loath to 
give up his views, and writes: **Comme il est facile de le voir 

les proteolytiques les plus puissants sont des bac- 

teries anaerobies stricts. Qe ne sont pas les seules, comme le 
croyait Bienstock; avec Martelly, nous avons vu qu'il y a des 
aerobes facultatifs capables de jouer le m^me r61e, mais leur 
action est loin d'etre aussi rapide." 

An organism to which Tissier and Martelly attached special 
significance is the facultative anaerobe, B. proteus vulgaris. In 

^ H. Tissier and Martelly: Ann, de l^Inst. Paslevr, xvi, p. 865, 1902. 
H. Tissier: ibid., xxvi, pp. 522-29, 1912. 
»E. Metchnikoff: ibid., xxii, p. 928, 1908. 

« Bienstock: Arch. /. Hyg., xxxvi, pp. 335-89, 1899; xxxix, p. 390, 1901. 
7L. F. Kettger: this Journal, ii, pp. 71-86, 1906-07; iv, pp. 45-55, 190S. 



J. A. Sperry and L. F. Rettger 455 

their study of the proteus group of bacteria, with special reference 
to its ability to initiate putrefactive changes, ttettger and Newell ^ 
employed twenty-five diflferent strains of proteus organisms, 
four of which came from the Pasteur Institute of Paris. In no 
instance could putrefaction be brought about through the agency 
of these organisms; nor was there any indication whatever of the 
decomposition of the special egg-meat medium by any member 
of this group when the tests were made under anaerobic con- 
ditions. Over half of the twenty-five strains of bacteria were 
those of proteus imlgarisy which has always been regarded as the 
most active proteolytic organism in the group. 

Representatives of the proteus group are frequently present 
in putrefactive mixtures, as Tissier and others showed; but, while 
they may aid materially in the ultimate decomposition of the 
intermediate products of putrefaction, their ability to initiate 
real putrefactive changes is far from being established. Ac- 
cording to Rettger the products of real putrefaction include 
hydrogen sulphide, mercaptan, and aromatic oxyacids, besides 
the usual products of tryptic digestion or proteolysis; namely, 
leucine, tyrosine, and tryptophane. No mention is made I)}' 
Tissier of the presence of mercaptan in any of his so called putre- 
faction tests w4th members of the proteus group. 

While none of the strict aerobes or facultative anaerobes used 
in the present investigation were able to utilize the pure animal 
and vegetable proteins without the aid of some intermediary' 
agent, it seemed highly probable that the putrefactive anaerobes 
could do so. It should be stated, however, that in earlier investi- 
gations' of these bacteria, and their deportment towards normal 
egg-white, no evidence was obtained that these organisms could 
decompose the egg albumin or even develop in the egg-white. 
The egg-white even appeared to have an antiseptic, and in a 
measure bactericidal, action on the anaerobes. This property 
was undoubtedly not due to the egg albumin, but to other agen- 
cies in the egg. 

The anaerobes which were employed in the present investi- 
gation were the well known putrefactive organisms, B. putri fiats. 

8 L. F. Rettger and C. R. Newell: ibid., xiii, p. 342, 1912-13. 
^ L. F. Rettger and J. A. Sperry: Jour. Med. Research, xxvi, pp. 55-64, 
1912. 



45^ Benarior ci Bacteria Towards Proteins 



^n^ tfa^ B<>^ii£SKr p^rroctlW iXMEtfiryk. Tabes ^rrxi'iAizrsc 
«r^rikr ^it^/^HO^oA QK^^sm v<re hatted ^ SCfC. for & f?nr nnfm&?g, 
in fJFfkr tft 0Xf0A tfe ffi^ oxrfen. They were ti^ti izKCoL^ced 
imfa tfa^ %xsi^snk0: Uj be iznn»ciipkted. the cixk&^boqs auui^ &a- 
i^fr^AAt. ^oA tivr toi'^^ iiKiibttted for foor d&T3 at 37^. \\ dbe 
^yl of thk pf^A tbtnt vas fHoalhr markai pocmairtkici. as iixE- 
#at^ (/f th^ partial ffi^estioo of the sofid mednxm and by the 
^^lanu^^^rkftkr odor of fntrefaetion. Sbmi ^^toiT tubes were inoe- 
nhttfA irfib tiro or tfar»!^ loopfuOs of the mediom taken from the 
waury layer. The aipu' tubes were incubated anaerolMC for 
f hr^; to four ^]ay«. The resultant surface growths were employed 
for the tnocfjlation of the pure-protein test media. 

rk/iutioTM of t^n albumin and of edestin were inoculated with 
minute quantiti^ of the agar surface growth in such a way as 
to prevent the inatsief of any of the culture medium itself. Tubes 
tA plain bouilkm were inoculated at the same time for eontrob. 
Tfje egg albumin tests were made with twenty-three different 
tulies, twelve bouillon controls being used. The>' were examined 
at the end (A 5, 7, 10, and 16 days. In the experiments on edestin 
fen edestin tubes and four controls were employed, and the ex- 
aminations were made after intervals of 4, 14, and 19 days^ and 
'\ weeks. 

In every instance the control tubes gave marked e\idence of 
growth with t3rpical putrefactive odor. All the egg albumin and 
fffUrntin tubes remained perfectly clear and odorless. Micro- 
ncAtpiv. examinations of Gram-stained films prepared from the dif- 
ferent me^lia showed that the plain bouillon tubes contained large 
numU;rs of the putrefactive bacteria, while but few were present 
in the mounts marie from the albumin or edestin medium, although 
all the tulx^ had Ixjcn inoculated with as nearly the same number 
of organisms as was possible. Transplantations made from the 
different sets of tubes to egg-meat medium resulted in the char- 
actrjristic putrefactive decomposition in the time usually required. 

While quantitative coagulation tests were not made, tubes of 
the egg albumin and the edestin medium were heated as previously 

'• UftttKcr: this Journal, ii, pp. 71-86, 1906-07. 



J. A. Sperry and L. F. Rettger 457 

described, after definite periods of incubation, and the amount 
of coagulable protein compared with that originally present. No 
differences could be noted. 

The above results seemingly justify the conclusion that even the 
putrefactive anaerobes, B. pvirificus, B. anthracis symptomaticiy 
and B. edematis maligni, are in themselves unable to bring about 
decomposition of native proteins, particularly ^g albumin and 
edestin. In order to determine whether these organisms may play 
the r61e of indirectly bringing about proteolysis of such proteins 
in a medium which contains, in addition to the native proteins, 
organic nitrogen that is readily assimilated, the following experi- 
ments were carried out. 

Large test-tubes containing 10 to 20 cc. of egg albumin medium 
(solution of pure egg albumin and inorganic salts previously 
described), to which 0.5 per cent peptone had been added, were 
inoculated with the putrefactive anaerol^es, and incubated at 
37°C. for five to fourteen days. Coagulation tests were then 
made and the loss of coagulable protein was determined. The 
results of three different series of tests were as follows. 

Experiment A, The uninoculatcd tube of medium (control) yielded 
0.1+ gm. of coagulated protein, while two tubes which were inoculated 
with^. anihracia aymptomatici contained only 0.01 -f and 0.02 -f- gm. re- 
spectively. Aside from the loss of coagulable protein, there were other 
indications of putrefactive decomposition, particularly odor and dis- 
coloration of the medium. The tubes had been incubated five days. 

Experitnent B. Eleven tubes of albumin-peptone medium were em- 
ployed, of which two were left uninoculated, four were inoculated with 
B. anthracis symptomaticiy and five with B. edematis maligni. At the end 
of the incubation period, which was ten days, the two control tubes con- 
tained 0.0210 gm. and 0.0217 gm. of coagulable albumin, whereas in all the 
inoculated tubes the coagulable protein had been completely digested, 
thus making final filtration and weighing unnecessary. 

Experiment C. In this last series of tests, in which the tubes were kept 
at 37 °r. for fourteen days, eighteen different tubes were used, of which • 
two served as uninoculated controls, eight were inoculated with B. anthracis 
symptamatici, and the same nimiber with J?, edematis maligni. The amounts 
of coagulable proteins in the controls at the end of the period were 0.0684 
gm. and 0.0679 gm. No coagulum could be obtained in any of the remaining 
sixteen tubes. 

While the putrefactive anaerobes, Uke the aerobes and faculta- 
tive anaerobes previously studied, arc unable to initiate proteoly- 



456 Behavior of Bacteria Towards Proteins 

B. anihracis symptoniatici , and B, edematis fnaligm. The organ- 
isms were grown continuously in the egg-meat medium recom- 
mended by one*® of us. Anaerobiosis was effected by the Wright 
and the Buchner (pyrogallol) methods. Tubes containing 
sterile egg-meat medium were heated at 80°C. for a few minutes, 
in order to expel the free oxygen. They were then inoculated 
with the anaerobe to be investigated, the conditions made an- 
aerobic, and the tubes incubated for four dajrs at 37°C. At the 
end of this period there was usually marked putrefaction, as indi- 
cated by the partial digestion of the soHd medium and by the 
characteristic odor of putrefaction. Slant agar tubes were inoc- 
ulated with two or three loopfulls of the medium taken from the 
watery layer. The agar tubes were incubated (anaerobic) for 
three to four days. The resultant surface growths were employed 
for the inoculation of the pure-protein test media. 

Solutions of egg albumin and of edestin were inoculated with 
minute quantities of the agar surface growth in such a way as 
t o prevent the transfer of any of the culture medium itself. Tubes 
of plain bouillon were inoculated at the same time for controls. 
The egg albumin tests were made with twenty-three different 
tubes, twelve bouillon controls being used. They were examined 
at the end of 5, 7, 10, and 16 days. In the experiments on edestin 
ten edestin tubes and four controls were employed, and the ex- 
aminations were made after intervals of 4, 14, and 19 days, and 
3 weeks. 

In every instance the control tubes gave marked evidence of 
growth with typical putrefactive odor. All the egg albumin and 
edestin tubes remained perfectly clear and odorless. Micro- 
scopic examinations of Gram-stained films prepared from the dif- 
ferent media showed that the plain bouillon tubes contained large 
numbers of the putrefactive bacteria, while but few were present 
in the mounts made from the albumin or edestin medium, although 
all the tubes had been inoculated with as nearly the same number 
of organisms as was possible. Transplantations made from the 
different sets of tubes to egg-meat medium resulted in the char- 
acteristic putrefactive decomposition in the time usually required. 

While quantitative coagulation tests were not made, tubes of 
the egg albumin and the edestin medium were heated as previously 

'« Rettger: this Journal, ii, pp. 71-86, 1906-07. 



J. A. Sperry and L. F. Rettger 457 

described, after definite periods of incubation, and the amount 
of coagulablc protein compared with that originally present. No 
differences could be noted. 

The above results seemingly justify the conclusion that even the 
putrefactive anaerobes, B. putrificus, B. anihracis sy^npUmmtici, 
and B. edeniatis malignly are in themselves unable to bring about 
decomposition of native proteins, particularly ^g albumin and 
edestin. In order to determine whether these organisms may play 
the role of indirectly bringing about proteolysis of such proteins 
in a medium which contains, in addition to the native proteins, 
organic nitrogen that is readily assimilated, the following experi- 
ments were carried out. 

Large test-tubes containing 10 to 20 cc. of egg albumin medium 
(solution of pure egg albumin and inorganic salts previously 
described), to which 0.5 per cent peptone had been added, were 
inoculated with the putrefactive anaerobes, and incubated at 
37°C. for five to fourteen days. Coagulation tests were then 
made and the loss of coagulablc protein was determined. The 
results of three different series of t^sts were as follows. 

Experiment A, The uninoculatcd tube of medium (control) yielded 
0.1 -f- gm. of coagulated protein, while two tubes which were inoculated 
with^. anthracia sumptornatici contained only 0.01 -f and 0.02 -f- gm. re- 
spectively. Aside from the loss of coagulable protein, there were other 
indications of putrefactive decomposition, particularly odor and dis- 
coloration of the medium. The tubes had been incubated five days. 

Experiment B. Eleven tubes of albumin-peptone medium were em- 
I)Ioyed, of which two were left uninoculated, four were inoculated with 
B. anthracis symptomatici, and five with B. cdematis maligni. At the end 
of the incubation period, which was ten days, the two control tubes con- 
tained 0.0210 gm. and 0.0217 gm. of coagulable albumin, whereas in all the 
inoculated tubes the coagulable protein had been completely digested, 
thus making final filtration and weighing unnecessary. 

Experiment C. In this last series of tests, in which the tubes were kept 
at 37°(\ for fourteen days, eighteen different tubes were used, of which 
two served as uninoculatcd controls, eight were inoculated with B, anthraeia 
symptomatici, and the same number withB. edematis maligni. The amounts 
of coagulable proteins in the controls at the end of the period were 0.0684 
gm. and 0.0679 gm. No coagulum could be obtained in any of the remaining 
sixteen tubes. 

While the putrefactive anaerobes, hke the aerobes and faculta- 
tive anaero})es previously studied, are unable to initiate proteoly- 



458 Behavior of Bacteria Towards Proteins 

sis of a native protein in a medium which contains no organic or 
other available nitrogen aside from the protein itself, they may 
indirectly bring about complete decomposition with the aid of 
other substances, like peptone, which readily furnish the nitro- 
gen (and car})on as well) that is necessary'' for early bacterial 
development. It must be assumed that in this early decomposition 
enzymes arc produced in sufficient amounts to cause proteolysis 
or digestion of the native protein. This view is the only one that 
appears to be tenable in connection with the work of the so called 
analytic bacteria in nature. The bacterial decomposition of 
animal and vegetable protein matter, which is of so much economic 
importance, is made possible through the agency of an interme- 
diary, like the nitrogenous products of autolysis, or any other nitrog- 
enous substance which furnishes sufficient nitrogen for the early 
development of the proteolytic bacteria and for the production 
of the enzymes which in themselves bring about the proteolysis. 
It is indeed surprising that no proteolysis should take place 
in a medium which contains pure native protein as the only 
possible source of nitrogen, even though the experiments extend 
over long periods of time, and at least small numbers of bacteria 
are constantly present. It was to be anticipated that even a 
slight increase in the number of bacteria, which at times fook 
place in the early part of our individual experiments, would be 
accompanied by the production of proteolytic enzymes, however 
small the amount. The absence of proteolysis may be regarded 
as an indication that the enzymes which are most directly con- 
cerned in the ^'bacterial*' decomposition of these proteins differ 
from ordinary pepsin and trypsin in that they have no apparent 
action in very high dilution, whereas the action of pepsin and tryp- 
sin does not depend on definite concentration. 

GENERAL CONCLUSIONS. 

The results of the present investigation are in full accord with 
the observations of Bainbridge on the behavior of certain aerobes 
and facultative anaerobes towards pure animal proteins. Our 
experiments have shown, furthermore, that the inability of 
bacteria to attack and decompose native proteins is not limited 
to these two groups of organisms, but that even the well known and 



J. A. Sperry and L. F. Rettger 459 

extremely active putrefactive anaerobes are devoid of this prop- 
erty. They have demonstrated also that pure vegetable prot<?ins, 
as for example edestin, show the same resistance to the direct 
action of bacteria. 

Solutions of native proteins may undergo complete proteolysis, 
however, if they contain peptone or some other nitrogenous food 
material which readily furnishes the necessary nitrogen for 
bacterial development. In such instance^ the proteolysis of the 
native protein is the immediate result of the action of an enzyme 
which has been elaborated by the bacteria during the process of 
rapid multiplication. This multiplication is made possible by 
the nitrogen-containing material which is present along with 
the native protein. 

The resistance of native proteins to direct decomposition b}' 
bacteria is not due to any antiseptic properties of the proteins, 
but to a construction of the molecule which renders it relatively 
stable, the component parts being so firmly bound together 
that a strong cleavage-producing agent, as extreme heat, strong 
acids and alkalies, and enzymes, is required to change them so 
that bacteria may utilize their products for cell nutrition. 



Electric Organic Combustion Furnace 

Replaceable Unit System (Patented) 




Has met with instant approval by leading chemists 

The furnaces can be used t(^ether or separately; each is 
equipped with rheostat control, allowing rapid or slow heat- 
ing or cooling. 

Upper half of furnaces can be raised while in use, exposing 
the combustion tube for observation purposes. 

Price* complete as Illustrated, $100.00 net 

Write today tor Illustrated folder to 



EIMER 



AND 



AMEND 



Founded 1851 
Headquarters for Educational and Industrial Laboratory Supplies 



NEW YORK CITY 




PITTSBURGH, PA. 



CONTENTS 

REriJEN L. Hill: Note on the use of colloidal iron in the determination of 
lactcise in milk 175 

David Fhaser Hauhis and IIenuy Jehmain Matde Crekjuton: Spectro- 
scopic invest igaticm of the reduction of hemoglobin by tissue reductane. 179 

Frank P. I'ndfhhill and Alhert CJ. Hogan: Studies in carbohydrate me- 
tabolism. V'lII. The inlluence of hydrazim* on the utilization of dextrose 203 

Frank P. Underhill and Albert (J. IIogan: Studies in carl)ohydrate me- 
tabolism. IX. The influence oi hydrazine on the nlvoxalase activity 
oUhe liver ". ". 211 

VicioR John Uardinq and Reginald M. MacLean: A colorimetric method 
for the estimation of amino-acid a-nitrogen 217 

Francis G. Henedict and Paul Roth: The metabolism of vegetarians as 
compared with the metabolism of non-vegetarians of like weight and 
height 231 

Francis G. Benedict and H. Monmouth Smith: The metabolism of athletes 
as compared with normal individuals of similar height and weight 243 

Francis G. Henedict and L. E. Emmes: A comparison of the basal metabo- 
lism of normal men and women 253 

Francis G. Benedict: Factors afTecting basal metabolism 263 

Francis G. Benedict: A respiration apparatus for small animals 301 

W. T. Bovie: Simple quarts mercury-vapor lamps for biological and photo- 
chemical investigations 315 

N. W. Janney: The metabolic relationship of the proteins to glucose 321 

Thomas B. Osborne and L.\fayette B. Mendel: The comparative nutritive 
value of certain proteins in growth, and the problem of the protein 
minimum 351 

Tho-mas B. Osborne and Lafayette B. Mendel: Further observations of 
the influenc!e of natural fats upon growth 379 

Victor C. Myers and Morris S. Fine: The non-protein nitrogenous com- 
pounds of the blood in nephritis, with special reference to creatinine 
ami uric acid 391 

Mary Louise Foster: Studies on a method for the quantitative estimation 
of certain groups in phospholipins 403 

IC. V. ^lcC■oLLUM and Marguerite Davis: The influence of the plane of 
protein intake on growth 415 

P. A. Levene and F. B. La Forge: On the mutarotation of phenylosazones 
of pentoses and hexoses 429 

P. A. Levene and F. B. La Furoe: On chondroitin sulphuric acid. Fourth 
paper 433 

Joel A. Sperry and' Leo F. Rettoer: The behavior of bacteria towards 
purified animal and vegetable proteins 445 

Robert M. Chapin and Wilmeu C. Powick: Correction. An imi>roved 
method for the estimation of inorganic phosphoric acid in certain tis- 
sues and food products 461 

The Journal of Biological Chemistry is designed for the prompt publica- 
tion of original investigations of a chemical nature in the biological sciences. 
It is issued monthly. Each volume consists of as many numbers as are required 
to make a total of between five and six hundred pages, m 

The price of The Journal to subscribers in the United States and Canada 
is S3.00 ptT volume; in other countries, $3.25. Remittances should be made by 
draft or check on New York or by postal money order, payable to The Journal 
OF Biological Chemistry, and should be sent to 2419-21 Grcemnount Avenue, 
Baltimore, Md., or to The Rockefeller Institute for Medical Research, 66th 
Street and Avenue A, New York City. 

Dr. Dakin is at present abroad. Manuscripts may be sent to any other of 
the editors, (-ontributors will receive 100 copi(*s of their papers free of cliarge, 
and additional copies will be furnished at cost. It is understood that, unless 
expressly agreed to the contrary, all manuscripts submitted to The Journal op 
Biological Chemistry shall not be offered for prior or shnultaneous publica- 
tion elsewhere. 

The price of this number is S2.00. 

THI WAVERLY PRESS 

SALTrMORE. U. 8. A. 




LIB 

VOL. XX APRIL, 1915 "^'^^ 4 ;|5 Nq. 4 

THE JOUENAL 



OP 

BIOLOGICAL CHEMISTRY 

FOUNDED BT CSmiaTlAN A. BGRTCB AND SUITAKNBO IN PAST BT TRB CBBIBTIAN A. BBBTBB 

ySlfOBIAL FUND 

EDITED BY 

H. D. DAKIN, New York City. LAFAYETTE B. MENDEL, New Haven, Conn. 
E. K. DUNHAM, New York City. A. N. RICHARDS, Philadelphia, Pa. 

DONALD D. VAN SLYKE, New York City. 

WITH THE COLLABORATION OP 

J. J. ABBL, BaltimoTB. Md. J. B. LEATHES, Toronto, Canada. 

R. H. CHITTBNDEN, New Hafen, Conn. P. A. LEVENB, New York. 

OTTO FOUN. Boaton, Masa. JACQUES LOEB, New York. 

WILUAM J. GIES, New York. A. S. LOEVENHART, Madison, Wla. 

L. J. HENDERSON, Cambridge, Hasa. GRAHAM LUSK, New York. 

RBID HUNT, Boston, Mass. A. B. MACALLUM, Toronto, Canada. 

W. A. JACOBS, New York. J. J. R. MACLEOD, Cleteland, Ohio 

WALTER JONES. Baltimore. Md. JOHN A. MANDEL, New York. 

J. H. KASTLB, LezinctOB. Ky. A. P. MATHEWS. Chicago. 01. 

F. G. NOVY. Ann Arbor, Mich. 

THOMAS B. OSBORNE, New Hafen, Conn. 

T. BRAILSFORD ROBERTSON, Berkeley, Cal. 

P. A. SHAFFER, St. Louis, Mo. 

A. B. TAYLOR, Philadelphia, Pa. 

F. P. UNDERHILL, New Haten, Conn. 

V. C. VAUGHAN, Ann Arbor. lUch. 

ALFRED J. WAKEMAN, New Haren, Conn. 

PUBUSHED MONTHLY 
BY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH 

FOR THE JOURNAL OF BIOLOGICAL CHEMISTRY, INC. 

2419^1 GREENMOUNT AVENUE, BALTIMORE. MD. 

Bntored aa aeoond-ebss matter, August 1, IQll. at the Post OfTice at Baltimore, Md., under the 

Aet of Hareh 8, 1870. 



CONCERNING THE ORGANIC PHOSPHORIC ACID 
COMPOUND OF WHEAT BRAN. IV. 

THE OCCURRENCE OF INOSITE TRIPHOSPHATE IN WHEAT 

BRAN. 

ELEVENTH PAPER ON PHYTIN. 

By R. J. ANDERSON. 

{From the Chemical Laboratory of the New York Agricultural Experiment 

Station, Geneva, and the First Chemical Institute oj the 

University of Berlin, Berlin.) 

(Received for publication, November 7, 1914.) 

INTRODUCTION. 

The only definitely homogeneous organic phosphoric acid ever 
isolated from wheat bran, as far as we are aware, is the crystalline 
inosite monophosphate which we described in a previous paper.^ 
This substance differs from other inosite phosphoric acids in that 
it crystallizes readily, and particularly in that its barium salt is 
very soluble in water; i.e., it is not precipitable with barium hy- 
droxide. 

Wheat bran, however, contains other organic phosphoric acids 
which are precipitated with barium hydroxide. It is evident from 
the work which we last reported on the subject* that this water- 
insoluble barium salt is not a homogeneous substance, but that 
it is a mixture of various organic phosphoric acids. 

Patten and Hart* isolated an acid from this mixture of the water- 
insoluble barium salts, which they believed to be identical with 
phytic acid, or the "anhydrooxymethylene diphosphoric acid'' 
of Postemak. 

^ R. J. Anderson: this Journal, xviii, p. 441, 1914; New York Agricul' 
tural Experiment Station Technical Bulletins, No. 36, 1914. 

2 Anderson: this Journal^ xviii, p. 425, 1914; New York Agricultural 
Experiment Station Technical Bulletins, No. 36, 1914. 

3 A. J. Patten and E. B. Hart: Am. Chem. Jour., xxxi, p. 564, 1904. 

463 

THE JOURNAL OF BIOLOGICAL CHIMIBTBT. VOL. ZX, NO. 4 



464 Inosite Triphosphate in Wheat Bran 

We have shown,* however, by the analyses of numerous prepara- 
tions freed from inorganic phosphate, that the above substance 
differs entirely in composition from phytic acid or salts of this 
acid. Although we were able to separate this substance into va- 
rious fractions which all differed in composition, it was impossible 
to obtain definitely homogeneous compounds by the method em- 
ployed. The separation was the more difficult since neither the 
barium salt nor other salts with inorganic bases crystallized. 

In our previous work on wheat brati we had always used the 
crude wheat bran ordinarily sold in the market for cattle feed. 
This product was not a pure bran, but contained various impurities. 
It is not unlikely that these impurities also contained organic 
phosphoric acids which would render the separation of pure com- 
pounds from the .final mixture more difficult. 

For the work reported in this paper we have used a perfectly 
pure wheat bran which was especially prepared from winter wheat 
for us in a local mill. 

From this pure bran the organic phosphoric acids were isolated 
as barium salts in the usual way (compare experimental part). 
The acid barium salt finally obtained was shaken up with cold 
water in which about one-half of the total substance dissolved. 

The examination of the water-insoluble portion forms the sub- 
ject of this paper. We hope later on to be able to separate the 
water-soluble substance also into its constituents. 

The portion insoluble in wat^r, as was shown in our last repoit,* 
contains a higher pyercentage of phosphorus and a lower pyercent- 
age of carbon than the water-soluble substance. 

Since the barium salt of the above organic phosphoric acid, as 
well as salts with other morganic bases, does not crystaUize, we 
tried to prepare salts with organic bases in the hope of obtaining 
crystalline compounds. It was found, however, that salts with 
phenylhydrazine, hydrazine hydrate, aniline, pyridine, etc., showed 
no more tendency to crystallize than those with inorganic bases; 
und the brucine salt is too soluble to crj'-stallize well. 

* .Anderson: this Journal, xii, p. 447, 1912; xviii, p. 425, 1914; New York 
Agricultural Experiment Station Technical Bulletins, No. 22, 1912, and 
No. 36, 1914. 

^ Anderson: this Journal, xviii, p. 425, 1914; New York Agricultural 
Experiment Station Technical Bulletins, No. 36, 1914. 



R. J. Anderson 465 

It was found finally that strychnine gave a readily crystalline 
salt with the above acid. This strychnine salt separates from the 
hot aqueous solution on cooling in long needle-shaped crystals, or 
t>onietimes in large thin plates, depending upon the concentration. 
The salt was recrj''stallized several times from wat€r. 

The substance thus purified had practically the same composi- 
tion and melting point as the salts described by Clarke,* which he 
had prepared from ^vild Indian mustard. 

However, it must be not^d that the molecular weight of such 
strychnine salts is so high that it is impossible accurately to de- 
termine the composition of the acid from the analysis of such com- 
pounds. We have used the strychnine salts, therefore, merely as 
a means of purification. After the salt had been several times 
recTvstallized, the strychnine was removed and the acid again 
precipitated with barium hydroxide. 

The substance was then analyzed; and the composition agreed 
\'(Ty closely with a neutral barium salt of inosit^ triphosphate, 
CfiHgOisPsBas. An acid barium salt was prepared by precipi- 
tuting the hydrochloric acid solution of the neutral salt with alco- 
hol. On analysis results were obtained which agreed with a com- 
pound of the following composition: CisHjsOttPQBas, which 
apparently represents three molecules of inosite triphosphate 
joined by five atoms of barimn. 

The free acid was prepared from the barium salt and analyzed. 
The results agreed closely with the calculated composition of inosite 
tripliosphate, CeHisOwPa. 

rnfortunately the barium salts were obtained only as amorphous 
or granular ix)wders, and the free acid itself was a non-crystalliz- 
ublc syrup. We believe, nevei-theless, that the substance repre- 

■ 

sonts a pure compound because it had been separated previously 
as a crystalline strychnine salt which appeared perfectly homo- 
jreiioous. 

< )n cleavage with dilute sulphuric acid in a sealed tube at a teni- 
jMiature of 150°, inosite triphosphate decomposes into inosite and 
]>h()sphori(' acid. 

'(J. Clarke: Jour, Chem, Soc^ cv, p. 535, 1911. 



466 Inosite Triphosphate in Wheat Bran 

EXPERIMENTAL PART. 

Isolation of the substance from wheat bran as a barium salt. 

The wheat bran was digested over night in 0.2 per cent hydro- 
chloric acid. It was then strained through cheese cloth and fil- 
tered. The free acid in the extract was nearly neutralized by add- 
ing a dilute solution of barium hydroxide until a slight permanent 
precipitate remained. A concentrated solution of barium chloride 
was then added and the precipitate allowed to settle over night. 
The supernatant liquid was syphoned off and the residue centrif- 
ugalized. The precipitate was finally brought upon a Biichner 
funnel, freed as much as possible from the mother-liquor, and 
then washed with 30 per cent alcohol. For further purification 
the substance was precipitated alternately six times from about 
1 pyer cent hydrochloric acid with barium hydroxide (Kahlbaum, 
alkali-free), and six times from the same strength hydrochloric 
acid with alcohol. 

After this treatment the substance was obtained as a snow-white 
amorphous powder. It was free from inorganic phosphate, and 
bases other than barium could not be detected. But it still con- 
tained oxalates from which it was freed in the manner described 
in a former paper.' The substance was finally precipitated from 
about 1 per cent hydrochloric acid with alcohol, filtered, and 
washed free of chlorides with dilute alcohol. It was finally washed 
with alcohol and ether and dried in vacuum over sulphuric acid. 
The dry substance weighed about 66 grams. 

This crude acid salt was rubbed up in a mortar with about 
500 cc. of cold water and allowed to stand a few hours. It was 
then filtered and washed successively with water, alcohol, and 
ether, and was dried in vacuum over sulphuric acid. The water- 
insoluble residue weighed about 30 grams. 

The water-soluble portion in the filtrate was precipitated with 
barium hydroxide and reserved for a future investigation. 



'Anderson: this Journal, xviii, p. 425, 1914; New York Agricultural 
Experiment Station Technical Bulletins^ No. 3G, 1914. 



R. J. Anderson 467 

Preparation of the strychnine salt. 

The water-insoluble portion of the barium salt mentioned above 
was suspyended in water and the barium precipitated with slight 
excess of dilute sulphuric acid. The barium sulphate wajs removed 
and the filtrate precipitated with excess of copper acetate. The 
copper precipitate was filtered and washed with water until the 
washings gave no reaction with barium chloride. It was then 
suspended in water, and the copper was precipitated with hydro- 
gen sulphide and filtered off. 

The solution containing the free organic phosphoric acid was 
diluted to about 2 liters with water, heated on the water bath, and 
44 grams of powdered strychnine were added. In a few minutes 
the strychnine was dissolved. The solution was filtered and con- 
centrated in vacuum at a temperature of 45°-50° to about one- 
half the volume. The strychnine salt soon began to separate in 
long needle-shaped crystals. After standing in the ice chest over 
night, the crystals were filtered off, washed in ice cold water fol- 
lowed by absolute alcohol and ether, and dried in the air. Yield, 
45.8 grams. 

For further purification the substance was twice recrystallized 
from hot water. It was then obtained in pure white needle- 
shaj)ed crystals which looked perfectly homogeneous. 

From more concentrated solutions the substance sometimes 
separates in the form of colorless plates which differ from the 
needle-shaped crystals in that they contain about 4 per cent more 
of water of crystallization. 

The strychnine salt has no sharp or definite melting point. 
Heated in a capillary tube it softens at about 200°C., but it does 
not melt completely even at a much higher temperature. On 
moist litmus paper it shows an acid reaction. The substance 
does not change in color on drying at 105°C. 

After drying at 105°C. in vacuum over phosphorus pentoxide, 
it was analyzed. For combustion the substance was mixed with 
fine copper oxide. 

The needle-shaped crystals gave the following: 

0.1534 gram substance gave 0.0901 gram H2O and 0.3329 gram COj. 

0.2944 gram substance gave 0.0702 gram MgaPjO:. 

0.1175 gram substance gave 6.75 cc. N at 22*C. and 763 mm. 



468 Inosite Triphosphate in Wheat Bran 

0.1688 gram substance lost 0.0154 gram HjO on dryinjr. 
0.1982 gram substance lost 0.0183 gram HjO on drying. 
Found: C=59.18; H=6.57; P=6.64; N=6.56; H20=9.12 and 9.23 per 
cent. 

The plate-shaped crystals gave the following: 

0.1786 gram substance gave 0.0993 gram HsO and 0.3909 gram CO2. 

0.4972 gram substance gave 0.1089 gram MgaPsO?. 

0.1955 gram substance lost 0.0256 gram HsO on drying. 

0.5997 gram substance lost 0.0783 gram HsO on drying. 

Found: C=59.60; H=6.22; P=6.10; H,O=13.09 and 13.05 per cent. 

These compounds apparently do not represent definite strych- 
nine salts of inosite triphosphate; but they seem to be mixtures 
of the tri- and tetrastrychnine salts. 

For tristrychnine inosite triphosphate, C»HuOuPj(CsiHs2N20s)8 = 1422. 
Calculated: C-58.23; H=5.70; P=6.54; N=5.90per cent. 
For tetrastrychnine inosite triphosphate, C»HuOiiPt(CsiHjiNsOi)<= 1756. 
Calculated: C=61.50; H=5.86; P=5.30; N= 6. 37 per cent. 

Preparation of the barium salt from the strychnine salt. 

The recrystallized strjxhnine salt, 27 grams, was dissolved in 
about 750 cc. of hot water, and the solution rendered alkaline with 
ammonia. After standing in ice water for some time the strych- 
nine was filtered off. The filtrate was shaken with several portions 
of chloroform to remove the last traces of strychnine. 

The solution, which now contained the ammonium salt of the 
organic phosphoric acid, was precipitated by adding a solution of 
barium chloride in excess. After settling over night, the precipi- 
tate was filtered and washed several times with water. It was 
then dissolved in 1 per cent hydrochloric acid, filtered, and 
precipitated with barium hydroxide in excess. The precipitate 
was filtered and washed vnth water until free from chlorides. 
It was again dissolved in 1 per cent hydrochloric acid, filtered, and 
the solution rendered neutral to litmus with barium hydroxide. 
The precipitate was filtered and washed free of chlorides with 
water and then in alcohol and ether and dried in vacuum over 
sulphuric acid. The substance was a pure white amorphous pow- 
der. On moist litmus paper it showed a ver}' faint acid reaction. 
It was free from nitrogen. 



R. J. Anderson 469 

For analysis it was dried at 100® in vacuum over phosphorus 
pentoxide. 

0.2323 gram substance gave 0.0370 gram HsO and 0.0758 gram CO2. 

. 2085 gram substance gave . 1668 gram B'aSOi and . 0868 gram MgsP207. 

Found: C-8.89; H=1.78; P=11.60; Ba=47.07 per cent. 

The substance was again dissolved in 1 per cent hydrochloric 
acid and the solution neutralized to litmus with barium hydroxide. 
The precipitated substance was filtered, washed free of chlorides 
with water, then washed with alcohol and ether, and dried in 
vacuum over sulphuric acid. 

It was analyzed after drying as before, and the following results 
were obtained: 

0.3168 gram substance gave 0.0481 gram HsO and 0.0972 gram GOs. 

. 2366 gram substance gave . 1968 gram BaSO^ and . 0960 gram MgsPiOr. 

Found: C=8.36; H=1.69; P-il.31; Ba-48.94 per cent. 

For the neutral barium salt of inosite trii^osphate, CcHfOiiP|Bai>>>826. 

Calculated: C-8.71; H=1.08; P=11.25; Ba-49.88 per cent. 

In the two analyses reported above the carbon is somewhat low. 
It must be noted, however, that these barium salts burned with 
extreme difficulty. Traces of carbon remained after prolonged 
heating in a current of oxygen. When the residues were mixed 
with fine copper oxide and rebumed, 1 or 2 mgm. more of carbon 
dioxide were obtained; but we believe that the combustion, even 
under these conditions, was incomplete. 

Preparation of the add barium salt. 

The above neutral barium salt was dissolved in the minimum 
quantity of 1 per cent hydrochloric acid; the solution was filtered 
and precipitated by adding about an equal volume of alcohoL 
The resulting precipitate was filtered, washed free of chlorides 
with dilute alcohol, washed further with absolute alcohol and 
ether, and dried in vacuum over sulphuric acid. The substance 
was then a pure white amorphous powder which showed a strong 
acid reaction on moist litmus paper. It was analyzed after dr>'- 
ing at 105°C. in vacuum over phosphorus pentoxide. 

0.2588 gram substance gave 0.0538 gram H3O and 0.1046 gram CO2. 
0.1691 gram substance gave 0.1027 gram BaS04 and 0.0890 MgaPaO;. 
Found: C = 11.02; H=2.32; P=14.67; Ba=35.74 percent. 



470 Inosite Triphosphate in Wheat Bran 

In the combustion of this substance a practically white ash was 
obtained. 

This compound is evidently a complex acid salt of inosite tri- 
phosphate and agrees with the following formula: 

C\8H350,5P9Ba4 = 1937, 

(calculated: C = 11.15; H = 1.80; P= 14.40; Ba=35.46 per cent), 
which may be graphically represented as follows: 

C8H,20i6P«=Ba 

>Ba 
C«HnOuP,=Ba 

>Ba 
CftHijOuPi^Ba 

That is, three acid molecules of inosite triphosphate joined by 
five atoms of barium. Whether it is a compound as represented 
above or a mixture of various acid salts of inosite triphosphate 
can hardly be determined. 

Preparation^}/ the free acid. 

The barium salt from above was suspended in water and the 
barium precipitated with a slight excess of dilute sulphuric acid. 
The barium sulphate was filtered off and the filtrate precipitated 
with excess of copper acetate. The copper precipitate was fil- 
tered and washed with water until it gave no reaction with barium 
chloride. It was then suspended in water and decomposed with 
hydrogen sulphide. The copper sulphide was filtered off, the fil- 
trate evaporated in vacuum at a temperature of 40°-45°C. to 
small bulk, and the residue dried in vacuum over sulphuric acid. 
The substance was then obtained as a practically colorless syrup. 
After continued drying it forms a hard, sticky, hygroscopic mass. 
It is extremely soluble in water, and also readily soluble in dilute 
and absolute alcohol. Much time was consumed in endeavoring 
to obtain it in crystalline form, b>it without success. The s3Tupy 
substance was therefore analyzed after drying: first, for several 
days in vacuum over sulphuric acid at room temperature; and 
finally at 100°C. in vacuum over phosphorus pentoxide. On 
drying at this temperature the color turned quite dark. 



R. J. Anderson . 471 

0.1685 gram substance gave 0.0a57 gram U2O and 0.1054 gram CO2. 
0.1693 gram substance gave 0.1363 gram MgsP207. 
Found: C = 17.06; H=3.69; P= 22. 44 per cent. 
For inosite triphosphate, G8Hi6Oi5Pj=420. 
Calculated: C = 17.14; H = 3.57; P = 22.14 per cent. 

Ill the above combustion a slight residue of unbumed carbon 
remained enclosed in the fused metaphosphoric acid. It was 
mixed with some fine coppyer oxide and again burned, when a few 
additional milligrams of carbon dioxide were obtained. 

Properties of inosite triphosphate. 

The reactions of this acid differ in several particulars from 
phytic acid or inosite hexaphosphate. 

The concentrated aqueous solution gives no precipitate with 
ammonium molybdate either in the cold or on warming. The cold 
aqueous solution of inosite hexaphosphate gives a white crystal- 
line precipitate with anunonium molybdate. 

The aqueous solution of the acid is not precipitated with silver 
nitrate, while inosite hexaphosphate gives a white amorphous 
precipitate. However, a solution of inosite triphosphate neutral- 
ized with ammonia gives a white amorphous precipitate with sil- 
ver nitrate. 

An aqueous solution of inosite triphosphate when added to a. 
solution of egg albumin causes only a slight turbidity. On longer 
standing a white precipitate separates slowly. Inosite hexaphos- 
phate precipitates egg albumin immediately. 

The acid is very soluble in water, and it is readily soluble in 
both dilute and absolute alcohol. From the latter solution it is 
precipitated by ether in small oily drops. 

The acid is not precipitated by barium or calcium chlorides; 
but alcohol added to the water solutions of acid and chlorides 
produces white amorphous precipitates. The calcium and barium 
salts are likewise precipitated with ammonia. 

Cleavage of inosite triphosphate into inosite and phosphoric acid. 

One gram of the acid, dissolved in a little water, was heated with 
10 cc. of N sulphuric acid in a sealed tube for three hours at 
150°-155°. The contents of the tube were then slightly yellowish 



472 loosite Triphosphate in Wheat Bran 

brown in color. The sulphuric and phosphoric acids were pre- 
cipitated with barium hydroxide and the inosite isolated in the 
usual way. Unfortunately a portion of the solution was lost, 
but from what remained 0.15 of a gram of inosite was obtained. 
After twice recrystallizing from dilute alcohol with the addition 
of ether 0.12 of a gram of inosite in the characteristic needle-shaped 
crystals was obtained. It gave the reaction of Scherer and melted 
at 222°C. (uncorrected). After mixing with spme previously 
isolated and analyzed inosite, the melting point did not change. 
The substance was therefore undoubtedly inosite, and the analysis 
was omitted. 

Although the barium salts described m this paper were amor- 
phous and the inosite triphosphate itself was a non-crystallizable 
syrup, we believe that the substances were pure. The basis for 
belief rests upon the fact that they had been prepared from the 
recrystallized strychnine salt, which appeared perfectly homo- 
geneous. 

We have been unable to complete the investigation of the water- 
soluble portion of the barium salt mentioned on page 466. We 
wish, however, to record the following data, although they are 
very incomplete. 

After the aqueous solution had been precipitated with barium 
hydroxide, as already mentioned, the precipitate was filtered and 
washed in water. It was then dissolved in the minimum quantity 
of about 2 per cent hydrochloric acid, filtered, and precipitated 
with about an equal volume of alcohol. The voluminous amor- 
phous precipitate was filtered and washed free of chlorides with dilute 
alcohol, washed with absolute alcohol and ether, and dried in 
vacuum over sulphuric acid. The dry substance weighed 38 
grams. 

It was again rubbed up in a mortar with about 200 cc. of cold 
water. After it had stood a short while, it was filtered, washed 
with water, alcohol, and ether, and dried ip vacuum over sul- 
phuric acid. The water-insoluble portion weighed 10.3 grams. 
The filtrate was concentrated in vacuum at a temperature of 40° 
to about 100 cc. As the concentration proceeded, a small quan- 
tity of a heavy white barium salt separated. It was not definitely 
crystalline and it weighed only 0.9 of a gram. The aqueous solu- 
tion, about 100 cc, contained, therefore, about 27 grams of the 



R. J. Anderson 473 

original substance. It was found impossible to obtain anything 
en'stalline from this solution. However, on heating to boiling, 
a heavy, semierystalline or granular powder separated. This 
was filtered and washed in hot water, alcohol, and ether, and dried 
in the air. It weighed 1.7 grams. 

The filtrate from the above was precipitated by adding about 
half a volume of alcohol. The amorphous w^hite precipitate was 
filtered, washed, and dried in vacuum over sulphuric acid. 

A further quantity of substance was precipitated by adding 
more alcohol to the filtrate. 

These three fractions were analyzed after drying at 105*^ in 
vacuum over phosphorus pentoxide. 

The semierystalline or granular powder gave the following: 

C = 10.14; n = 1.90; P=13.18; Ba=39,60; H20=12.n i^r cent. 

These results agree closely with the figures for a dibarium salt 
of inosite triphosphate. 

For CgHiiOisPsBa,. 

Calculated: C=10.41; H = 1.59; P=13.45; Ba= 39. 79 per cent. 

1^he first amorphous precipitate gave: 

r = 12.07; H=2.30; P=13.93; Ba=33.70 per cent. 

The second amorphous precipitate gave: 

C = 13.54; H = 2.74; P=13.22; Ba= 32. 38 per cent. 

Lack of time has prevented the further examination of these 
substances, but we hope to complete the investigation lat<*r. 

From the results recorded it is evident that the organic phos- 
phoric acid contained in the preparation used in this investiga- 
tion consisted principally of inosite triphosphate, CeHisOuPj. The 
amorphous barium salts analyzed above may have contained some 
inosite diphosphate or some other organic phosphoric acid. In 
addition to the above it must be noted that inosite monophosphate, 
described in an earlier paper, had been isolated from this same 
wheat bran. 

The author desires to express his appreciation and thanks to 
His Excellency, Prof. E. Fischer, for the interest shown in the work 
reported in this paper. 



THE HYDROLYSIS OF PHYTIN BY THE ENZYME 
PHYTASE CONTAINED IN WHEAT BRAN. 

TWELFTH PAPER ON PHYTIX. 

By R. J. ANDERSON. 

{From the Chemical Laboratory of the New York Agricultural Experiment 

Statioriy Geneva, and the Ludwig Mond Biochemical Research Laboratory 

of the Institute of Physiology, University College, London.) 

(Received for publication, January 9, 1915.) 
INTRODUCTION. 

It has been shown by the investigations of Suzuki, Yoshimura, 
and Takaishi^ that rice bran contains an enzyme which rapidly 
hydrolyses phytin with formation of inosite and inorganic phos- 
phoric acid. These authors concluded that wheat bran likewise 
contained a similar enzyane, because the inorganic phosphorus 
increased in wheat bran extracts on standing. 

Plimmer^ examined a large number of extracts prepared from 
the int<?stine8, liver, pancreas, castor beans, etc., as to their 
action on organic phosphorus compounds. While some of these 
showed a slight cleavage action on phytin, none could be com- 
pared in activity to an aqueous extract of wheat bran. The 
hydrolytic action of these extracts was determined by estimating 
from tune to time the amount of inorganic phosphorus split off 
from phytin solutions of known concentration. 

Since the above experiments clearly demonstrated that large 
quantities of inorganic phosphate were liberated from ph5rtin 
by wheat bran extracts, it appeared of interest to determine what 
products, in addition to inosite and inorganic phosphoric acid, 
were fonned under these conditions. For this purpose wheat 
bran extract was allowed to act upon a dilute solution of phytin 

^ U. Suzuki, K. Yoshimura, and M. Takaishi : Bulletins of the College of 
Agriculture, Tokyo, vii, pp. 503-512, 1907. 

* R. H. A. Plimmcr: Biochem. Jour., vii, p. 43, 1913. 

475 



476 Phytase in Wheat Bran 

at a temperature of 37*^. Inorganic phosphoric acid was deter- 
mined in the solution from time to time. 

It was found that about two-thirds of the total phosphorus was 
split off durmg the first sixteen days. Afterwards there was no 
appreciable change even on standing for about two years. 

The solution had been prepared and the original determinations 
made loy Dr. Plimmer. At his suggestion the writer undertook 
to examine the final reaction mixture for such products as had 
been formed. 

These products were separated into two portions by precipi- 
tating the original solution with barium hydroxide. The precipi- 
tate contained inorganic barium phosphate and also those barium 
salts of organic phosphoric acids that were insoluble in the dilute 
alkaline solution. The filtrate, on the other hand, was found to 
contain inosite monophosphate, and free inosite. 

The inorganic phosphate and other impurities were removed 
from the crude barium hydroxide precipitate, as will be described 
in the experimental part. The organic phosphoric acids which 
remained were obtained as amorphous barium salts. It was im- 
possible to isolate any unchanged barium phytate. It is evident 
then that all the phytin had been partially hydrolyzed. 

The above amorphous substance appeared to consist mainly 
of barium inosite triphosphate, but probably mixed with some 
barium inosite diphosphate. Owing to the difficulty of separating 
these compounds their isolation was not attempted. 

Among the soluble substances which had been formed we were 
able to isolate and identify inosite monophosphate, a substance 
which we have previously isolated from wheat bran.' In addi- 
tion to this, the solution also contained some free inosite which 
was isolated by means of its lead compound. 

The action of this enzjine, ph5rtase, upon phytin appears to 
proceed in several stages. Only a portion of the ph5rtin is com- 
pletely decomposed into inosite and phosphoric acid, but all the 
phj-tin is partially hydrolyzed with formation of certain lower 
phosphoric acid esters of inosite; viz,y inosite tri-, di-, and mono- 
phosphate, and hiorganic phosphoric acid. The formation of 
these intennediatc products is only possible through the destruc- 

' R. J. Anderson: this Journal, xviii, p. 441, 1914; and New York Agri' 
cultural Experiment Station Technical BulletinSf No. 36, 1914. 



R. J. Anderson 477 

tion or inhibition of the enzjine before the hydrolysis is complete. 
The reason for this inhibition is not clear, but it may be due to 
the excess of phosphoric acid which is liberated. 

It is interesting to note, and we call particular attention to 
the fact, that the organic phosphoric acids which remain as inter- 
mediate products of the action of the enzyme upon phytin, viz.y 
inosite triphosphate and iriosite monophosphate, are identical with 
the substances which we have isolated previously from wheat 
bran after it has been digested in 0.2 per cent hydrochloric acid. 

EXPERIMENTAL PART. 

Commercial phytin, 100 grams, was dissolved in 500 cc. of 
water and filtered from the insoluble matter which weighed 3.5 
grams when dried at lOO*'. The pale yellow solution was treated 
with 38 grams of oxalic acid dissolved in about 250 cc. of water. 
The calcium oxalate was filtered off, washed, and dried. It 
weighed 48 grams. The solution was diluted to 6000 cc. with 
water and was then found to contain 40 grams of phosphorus 
})entoxide. To it were added 800 cc. of an aqueous extract of 
wheat bran which contained 2.2 grams of P2OB. The solution 
was kept under toluene at a temperature of 37°. No hydrolysis 
occurred in a week. This was evidently due to the strongly acid 
reaction of the solution. It was nearly neutralized with ammonia, 
and 735 cc. of bran extract containing 1.53 grams of P205were 
added. It was again kept at a temperature of 37° under toluene. 
In nine days one-half of the total phosphorus was hydrolyzed; 
ill sixteen days two-thirds was hydrolyzed. In thirty-five days 
the amount of hydrolysis had not altered, and after about two 
\'ears it was again the same. The total and inorganic phosphorus 
was detennined as described by Plimmer and Page.^ 

The dark colored solution was filtered and barium hydroxide 
(Kahlbaum) added in slight excess. After standing over night 
th(* precipitate was filtered and washed in water. The filtrate 
and washings were evaporated on the water bath and the residue 
was examined, as will be described later. 

The barium precipitate was dissolved in about 2.5 per cent 
hydrochloric acid, filtered, and precipitated by adding about an 

* U. H. A. Plimmer and II. J. Page: Biochem. Jour., vii, p. 162, 1913. 



478 Phytase in Wheat Bran 

equal volume of alcohol. The precipitate was filtered and washed 
in dilute alcohol. The substance was again precipitated four times 
in the same way. It was then precipitated by barium hydroxide 
three times from about 2 per cent hydrochloric acid, and finally 
twice more with alcohol from the same strength hydrochloric 
acid. After finally filtering it was washed free of chlorides in 
dilute alcohol and then in alcohol and ether and dried m vacuum 
over sulphuric acid. The substance was then a snow-white amor- 
phous powder. It weighed 28.4 grams. It was free from chlo- 
rides and inorganic phosphate, and bases other than barium 
could not be detected. 

The substance was then rubbed up in a mortar with 300 cc. 
of cold water and allowed to stand with occasional shaking for 
a few hours. It was then filtered and washed in water, alcohol, 
and ether, and dried in vacuum over sulphuric acid. The drj^ 
water-insoluble portion weighed 5.4 grams. 

The filtrate from above was neutralized to litmus with barium 
hydroxide. The precipitate was filtered and washed in water, 
alcohol, and ether, and dried in vacuum over sulphuric acid. It 
weighed 23.6 grams. 

These precipitates were analyzed after drying at 105^ in vacuum 
over phosphorus pentoxide. 

The first, water-insoluble portion gave the following result : 

Found: C = 11.46; H = 1.93; P = 11.59; Ba = 39.94 per cent. 

This substance is apparently largely composed of the dil)arium 
inosite triphosphate. 

Calculated for this: CeHiiOisP.Baj = 690. 

C = 10.43; H = 1.59; P = 13.47; Ba = 39.71 per cent. 

It is, however, not pure, but apparently contauis some barium 
inosite diphosphate; because the carbon is high and the phosphorus 
is low. 

The water-soluble substance which was precipitated with bar- 
ium hydroxide gave the following: 

Found: C = 9.63; H = 1.63; P = 10.91; Ba = 47.41 per cent. 

This substance also appears to consist largely of the neutral 
barium salt of inosite triphosphate. 



R. J. Anderson 479 

Calculated for the latter: CcHjOuPsBa, = 826. 

C = 8.71; H = 1.08; P = 11.25; Ba = 49.88 per cent. 

The carbon, however, is high, and the phosphorus as well as 
the barium are low; which points to the presence of barium 
inosite diphosphate. 

In the hope of approximately separating these barium inosite 
tri- and diphosphates, the substance, 23.6 grams, was digested 
in dilute acetic acid for several hours with occasional shaking. 
It was then filtered and washed in water, and the insoluble por- 
tion dried in vacuum over sulphuric acid. It weighed 10 grams. 

The filtrate and washings containing the soluble portion of the 
substance were precipitated by adding lead acetate in excess. 
After standing over night the white amorphous precipitate was 
filtered and washed in water. It was suspended in water and 
decomposed by hydrogen sulphide, filtered, and the excess of 
hydrogen sulphide boiled off. It was again precipitated in the 
same manner with lead acetate and decomposed with hydrogen 
sulphide. The solution still contained a considerable quantity 
of barium. The barium was therefore removed with a slight 
excess of dilute sulphuric acid. After filtration of the barium 
sulphate the solution was precipitated by adding copper acetate 
in excess. The copper precipitate was filtered, washed, suspended 
in water, and decomposed with hydrogen sulphide. After remov- 
ing the copper sulphide, the filtrate was evaporated in vacuum to 
small bulk, and finally dried in vacuum over sulphuric acid. 
There remained a thick, nearly colorless syrup. It was readily 
soluble in alcohol. The addition of chloroform to this solution 
caused the substance to separate in small oily drops; the addition 
of ether produced a cloudiness, and on standing a flocculent amor- 
phous precipitate separated. These solutions could not be brought 
to crystallize. The acid preparation itself was kept for several 
weeks in the desiccator over sulphuric acid. It became a hard, 
sticky mass, but showed absolutely no tendency to crystallize. 
The color of the preparation, kept in this manner, gradually 
darkened. 

Since the acid would not crystallize, the syrupy substance 
was analyzed after dr\ing at 105° in vacuum over phosphorus 
pentoxide. 

• 
THE JOURNAL OF BIOLOGICAL CHEMISTUV, VOL. XX, NO. 4 



480 Phytase in Wheat Bran 

Found: C = 18.58; H = 3.82; P =- 20.38 per cent. 
Calculated for inosite triphosphoric acid: CiHiiOuPs = 420. 
C = 17.14; H = 3.67; P = 22.14 per cent. 
Calculated for inosite diphosphoric acid: CiHuOisPs = 340. 
C = 21.17; H = 4.11; P = 18.23 per cent. 

This acid preparation is evidently also a mixture of the inosite 
tri- and diphosphoric acids. 



EXAMINATION OF THE FILTRATE AFTER THE WATER-INSOLUBLE 

BARIUM SALTS HAD BEEN PRECIPITATED. 

The filtrate was evaporated, as mentioned on page 477, and the 
residue taken up in hot water. It was decolorized with animal 
charcoal. The solution was neutral in reaction. It strongly re- 
duced Fehling's solution on boiling, possibly due to sugars intro- 
duced with the bran extract. The solution was found to contain 
barium and also phosphorus in organic combination, evidently 
inosite monophosphate. The aqueous solution was precipitated 
by adding about an equal volume of alcohol and the white amor- 
phous precipitate filtered off, the filtrate being reserved for further 
examination. 

Isolation of inosite monophosphate. The above precipitate, which 
formed on the addition of alcohol, was dissolved in water, slightly 
acidified with acetic acid, and then precipitated with lead acetate 
in excess. After settling, this was filtered, washed in water, 
suspended in hot water, and decomposed with hydrogen sulphide. 
It was then filtered, and the filtrate boiled to expel excess of 
hydrogen sulphide. It was reprecipitated several times with lead 
acetate in the same manner until a white lead precipitate was 
obtained. This was finally decomposed with hydrogen sulphide, 
filtered, and evaporated to small bulk in vacuum and then dried 
in vacuum over sulphuric acid until a thick syrup remained. On 
scratching with a glass rod, this crystallized to a white solid mass. 
It was digested in alcohol and filtered, washed in alcohol and ether, 
and diied in the air. It weighed 1.6 grams. It had all the proper- 
ties of inosite monophosphate. For further purification it was dis- 
solved in a few cubic centimeters of water and filtered. Alcohol 
was then added until the solution turned cloudy; it was heated 



R. J. Anderson 481 

until it cleared up, and more alcohol was added until a faint 
permanent cloudiness remained. It was allowed to stand for 
about forty-eight hours at room temperature, when the substance 
had separated in massive, practically colorless crystals. After 
filtering, washing in alcohol and ether, and drying in the air, 
1 gram of substance was obtained. When heated in a capillary 
tube it began to soften at 188°-189° and melted with decomposi- 
tion and effervescence at 190°C. (uncorrected). The appearance 
and properties of the substance corresponded exactly with those 
described for inosite monophosphate and the analysis was there- 
fore omitted. 

Isolation of inosite. The filtrate, after precipitation of the 
above barium salt of inosite monophosphate with alcohol, was 
evaporated on the water bath until the alcohol was removed. 
It still contained barium, chlorides, etc. The barium was quan- 
titatively precipitated with dilute sulphuric acid and the solu- 
tion again concentrated on the water bath. 

The addition of lead acetate caused no precipitate. Basic lead 
acetate was then added as long as any precipitate formed. This 
precipitate was filtered off and discarded. The solution was 
then heated to boiling, more basic lead acetate added, and the 
solution was finally made strongly alkaline with ammonia and 
allowed to stand over night. This precipitate was filtered and 
washed in water, and then decomposed in aqueous suspension 
with hydrogen sulphide. The filtrate was concentrated on the 
water bath and the inosite brought to crystallization by the addi- 
tion of alcohol. After recrystallizing several times 0.5 of a gram 
of pure inosite was obtained in the characteristic needle-shaped 
crystals. It gave the reaction of Scherer and melted at 218*^0. 
(uncorrected). 

SUMMARY. 

The chief products of the hydrolysis of phytin by the pliytasc 
ill wheat bran are inorganic phosphoric acid and certain inter- 
mediate compounds apparently consisting of inosite tri-, di-, and 
monophosphates. These intermediate substances are identical 
with the compounds which we have previously isolated from 0.2 
per cent hydrochloric acid extracts of wheat bran. 



482 Phytase in Wheat Bran 

A portion of the phytin was completely hydrolyzed by the 
action of the enzyme into phosphoric acid and inosite; because 
the solution was found to contain some free inosite. 

All the ph>i:in was at least partially hydrolyzed; since the 
final reaction mixture did not contain any unchanged inosite 
hexaphosphat€. 



THE HYDROLYSIS OF THE ORGANIC PHOSPHORUS 

COMPOUND OF WHEAT BRAN BY THE 

ENZYME PHYTASE. 

THIRTEENTH PAPER OS PHYTIN. 

By R. J. ANDERSON. 

{From the Chemical Laboratory of the New Y'ork Agricultural Experiment 
Station, Geneva, and the Ludwig Mond Biochemical Research Labora- 
tory of the Institute of Physiology, University College, London.) 

(Received for publication, January 9, 1915.) 

IXTRODUCTION. 

In several previous papers^ on the subject of the organic phos- 
phoric acid compounds of wheat bran, we have mentioned that 
when wheat bran is digested in 0.2 per cent hydrochloric acid, 
tlie resulting extract always contains relatively large quantities 
of inorganic phosphate. Up to the present we have had no data 
concerning the origin of this inorganic phosphate, and it remained 
to determine whether it was originally present in the bran or 
had lx»en fonned during the digestion by hydrolysis from the 
organic phosphorus compound. 

Hart and Andrews,- who examined wheat bran, as well as a 
large number of other feeding materials, came to the conclusion 
that the phosphorus was present almost entirely in organic com- 
bination. Some criticism, however, has been offered of their 
method of estimating inorganic phosphorus, and it has l)pen sug- 
gested that the time, fifteen minutes, which they allowed for 
digestion was not sufficient for complete extraction.^ 

^ K. J. Anderson: this Journal, xii, p. 447, 1912; xviii, p. 425, 1914; New 
York Agricultural Experiment Station Technical Bulletinfi, No. 22, 1912 and 
No. 36, 1914. 

^ E. B.Hart and W. H. Andrews: New York Agricultural Experiment 
Station Bulletins, No. 238, 1903. 

' Forbes: Ohio Agricultural Experiment Station Bulletins, No. 215. 

483 



484 Organic Phosphorus Compound of Wheat Bran 

From the work of Suzuki, Yoshimura, and Takaishi* and of 
Pliinmer* it is evident that wheat bran contains an enzyme which 
rapidly hydrolyzes a portion of the organic phosphorus into inor- 
ganic phosphoric acid. The quantitative determinations of the 
activity of this enzyme on phytin or inosite hexaphosphate re- 
ported in the preceding paper show that about two-thirds of 
the total phosphorus of the phytin was transformed into inor- 
ganic phosphoric acid in about sixteen days. Since a small quan- 
tity of wheat bran extract is capable of hydrolyzing phytin to 
such an extent, it is evident that the organic phosphoric acid 
compound originally present in the bran would also be hydro- 
lyzed on digestion under the same conditions. 

The determinations herein reported were undertaken in the 
hope of throwing some light upon this question. It was also 
hoped to determine the cause of the large percentage of inorganic 
phosphoric acid in extracts obtained on digesting wheat bran 
in 0.2 per cent hydrochloric acid. 

The results show that the enzyme phytase contained in wheat 
bran very rapidly hydrolyzes the organic phosphorus compound 
of the bran into inorganic phosphoric acid under certain con- 
ditions. We have endeavoured to determine under what con- 
ditions the maximum activity of the enzyme manifests itself, 
and also the conditions under which its action is either inhibited 
or destroyed. 

When wheat bran is digested in distilled water the hydrolysis 
begins at once and proceeds with considerable rapidity, and at 
tlie end of twenty-four hours nearly 90 per cent of the total 
water-soluble phosphorus is inorganic. The organic and inor- 
ganic phosphates of the bran are only partially soluble in the 
water. When the extracts, after digestion of the bran for twenty- 
four hours in water, are acidified with hydrochloric acid, only 
about 60 per cent of the total dissolved phosphorus compounds 
are inorganic (see Table II). This result is not perceptibly altered 
by digesting the bran in water for forty-eight hours and then 
acidif\nng (sec Table III). 

* U. Suzuki, K. Yoshimura, and M. Takaishi: BullcttJis of the College of 
Agriculture, Tokyo, vii, pp. 503-512, 1907. 

^ R. H. A. Plimmer: Blovhem. Jour.j vii, p. 43, 1913. 



R. J. Anderson • 485 

When the bran is digested in 0.2 per cent hydrochloric acid, 
the rate of hydrolysis is greater than with pure water. The 
actual amounts of both organic and inorganic phosphorus are 
greater in the dilute acid than in the aqueous extracts. After 
three hours about one-third of the total phosphorus is inorganic, 
and after twenty-four hours about three-quarters of the total 
soluble phosphorus has been hydrolyzed to inorganic phosphoric 
acid, as is shown in Table V. 

With 0.2 per cent acetic acid the hydrolysis is more rapid and 
more complete than with 0.2 per cent hydrochloric acid, 85 per 
cent in twenty-four hours (Table VII); and the most complete 
hydrolysis was obtained with 0.1 per cent hydrochloric acid; 
namely, 92 per cent (Table VIII). 

That the hydrolysis of the organic phosphorus compound is 
due to enzyme action and not to the acid employed is strikingly 
illustrated by the fact that with increasing strengths of the hydro- 
chloric acid the amount of inorganic phosphorus rapidly and 
steadily diminishes. But when 0.5, 1.0, or 2.0 per cent hydro- 
chloric acid is used, the inorganic phosphorus is practically con- 
stant (Table V). 

The enzyme is remarkably sensitive to certain concentrations 
of hydrochloric acid. The hydrolyzing action is stimulated by 
the presence of 0.1 or 0.2 per cent hydrochloric acid, in which 
the action is much more rapid than in pure water; but 0.3 per 
cent almost inhibits the enzyme activity. With the 0.1 per cent 
acid 92 per cent, and with the 0.2 per cent acid 76 per cent, of 
the total phosphorus becomes inorganic, but with 0.3 per cent 
acid only about 20 per cent of the total phosphorus becomes 
inorganic (Tables V and VIII). 

Since such a slight difference in strength of the hydrochloric 
acid caused such a great difference in the amount of the inorganic 
phosphorus, it appeared of interest to determine whether the 
acid merely inhibited the hydrolysis, or if the enzyme was de- 
stroyed by it. Some of the bran was digested in 0.5 per cent 
hydrochloric acid for one and one-half hours, and the acid then 
nearly neutralized with ammonia, and the whole allowed to digest 
for twenty-four hours. In this case there was no increase in the 
inorganic phosphorus, showing that hydrochloric acid of this 
strength coiiiplctoly destroys the enzyme (Table X). 



486 Organic Phosphorus Compound of Wheat Bran 

The enzynie is also destroyed by digesting the bran in 0.25 
per cent ammonia, as well as by pouring boiling water or boiling 
0.2 per cent hydrochloric acid over the bran (Tables X and IX). 

Concerning the amount of inorganic phosphorus in wheat bran, 
it is evident, judging by the results obtained with 0.5, 1.0, and 
2.0 per cent hydrochloric acid, as well as on treating bran with 
boilmg water or boiling 0.2 per cent hydrochloric acid and with 0.25 
per cent ammonia, that ordinary bran contains about 0.1 per 
cent of phosphorus as inorganic, which represents about 11 per 
cent of the total soluble phosphorus. It seems quit<j certain that 
under these conditions of extraction no hydrolysis of the organic 
phosphorus occurs, and we believe that these figures represent 
the amount of inorganic phosphorus normally present in wheat 
bran. 

These results are not at variance with the findings of Hart 
and Andrews.* After digesting bran in 1.0 per cent hydrochloric 
acid for forty hours, they found 0.087 per cent of morganic 
phospliorus in the extract. This is practically the same as our 
result, 0.081 per cent inorganic phosphorus, after digesting bran 
for five hours in 1.0 per cent hydrochloric acid. The same authors 
found only 0.036 per cent inorganic phosphorus in wheat bran 
after digesting in 0.2 per cent hydrochloric acid for fifteen minutes. 
The lower percentage is no doubt due to incomplete extraction, 
and we believe that the higher figures represent the actual amount 
present. 

EXPERIMENTAL PART. 

Total and inorganic phosphorus was determined in extracts 
prepared from wheat bran. The extracts were prepared by di- 
gesting wheat bran in water or dilute acid for varying lengths of 
time, the particulars and details being given at the beginning of 
each table. Total phosphorus was detennined after decomposing 
by the Neumann method. Inorganic phosphorus was determined 
as follows: The extract was diluted with about 50 cc. of water, 
15 grams of ammonium nitrate were added, and then warmed 
to 65° on the water bath. It was then strongly acidified with nitric 
acid, the phosphorus precipitated with anmionium molybdate, 
and the solution kept at the above temperature for half an hour. 

^ Hart and Andrews: loc. cit. 



R, J, Anderson 487 

lender these conditions there is no danger of cleavage of the 
organic phosphorus, — at least weighable quantities of phosphorus 
are not precipitated during this time from preparations which 
are free from inorganic phosphate. Plimmer^ has also shown 
that it is necessary to heat phytin solutions with ^^ or -J nitric 
acid for several hours to a temperature of 75° or higher before 
appreciable quantities of inorganic phosphorus are split off. 

The phosphomolybdate precipitate was then filtered off and 
the phosphorus determined as maguesium pyrophosphate in the 
usual way. 

The digestions were made throughout these experiments at 
room temperature, about 16°C. Ten grams of the bran were 
digested in 100 cc. of water for the time mentioned, then filtered, 
and 20 cc. used for each determination. The same quantities 
were used throughout. 









TABLE I. 






T0TA.1. 


PHOSPHORUS 


INORGANIC PHOSPHORUS; 


TOTAL PHOSPHORUS 


'IME 


IN 


EXTKACT 

per cfnt 




' IS EXTRACT i 

per rent 


AS INORGANIC ' 


hrfi. 


per cent 


• 

i 


1 


0.362 




0.165 


45.77 


1 




0.420 




0.239 


56.95 


4 




0.508 




0.383 


75.34 


20 


1 


0.655 




0.576 


88.00 


24 


, 


0.747 




0.660 

1 


88.43 



It will be noticed that the hydrolysis is greatest during the first 
four hours. The percentage of phosphorus represents only that 
amount which was soluble in the water. 

After digesting in water for twenty-four hours and then acidi- 
fying with hydrochloric acid and shaking for half an hour the 
following results were obtained: 

TABLE H. 

I TOTAL PHOSPHORUS IN'OROANrc PHOSPHORUS TOTVL PHOSPHORIK 
"^^^^ IK EXTRACT IX EXTRACT ' \A INORGANIC 

Arff. per cent per cent per cent 

24 I 1.00 0.649 60.85 



' Plimmer: loc. di., p. 72. 



488 Organic Phosphorus Compound of Wheat Bran 



After digesting the bran for forty-eight hours in water and then 
adding 100 cc. of 2 per cent hydrochloric acid and allowing it 
to stand for another twenty-four hours, the following results were 
obtained : 





TABLE III. 




TOTAL PHOBPHORUS 


XNOBGANIC PHOSPHORCB 


TOTAL PHOSPHORUS 


IN EXTRACT 


m EXTRACT 


AS INORGANIC 


percent 


per cent 


per cent 


1.27 


0.803 


62.79 


1.28 


, 0.796 


62.13 



The addition of toluene to the water appears to diminish the 
amount of phosphorus dissolved without affecting the degree of 
hydrolysis. 



TIME 



hra. 

24 
with toluene 



TABLE IV. 



TOTAL PHOSPBORL^ 
IN EXTRACT 



per cent 

0.518 



INORGANIC PHOSPHORUS, 
IN EXTRilCT 



percent 

0.462 



TOTAL PHOSPBORrS 
AS INORGANIC 

per cent 

89.25 



With hydrochloric acid as the extracting medium, the strength 
varying from 0.2 to 2.0 per cent, the following results were ob- 
tained: 10 grams of bran were digested in 100 cc. of the acid for 
the length of time mentioned in the table. It was then filtered 
and 20 cc. of the filtrate were used for each determination. 







TABT*K V. 










1 TOTAL 


INORGANIC 


TOTAL 


SOLVENT 


TIME 


PHOSPHORUS IN 


! PHOSPHORUS IN 


PHOSPHORUS AS 




hrs. 


1 EXTRACT 


1 EXTRACT 


INORGANIC 




1 

percent 


percent 


per cent 


0.2% HCl 


3 


1.130 


0.415 


36.52 


0.2% HCl i 


24 


i 1.200 


1 0.925 


76.87 


0.3% HCl 


24 


0.999 


0.203 


20.22 


0.4% HCl 


24 


0.939 


0.146 


15.61 


0.5% HCl 


20 


0.922 


0.124 


13.44 


1.0% HCl 


5 


0.894 


: 0.081 


9.13 


2.0% HCl 


30 


1.080 


0.117 


10.83 



R, J. Anderson 



489 



The presence of toluene did not materially alter these results. 
Hydrochloric acid, 0.5 per cent, after twenty hours, with toluene 
gave the following: 

TABLE VI. 



TIME 

hrs. 

20 



TOTAL PHOSPHORUS INOBOAK IC PHOSPHORUS ' TOTAL PHOSPHORUS 
IN EXTRACT IK EXTRACT AS INOROAKIC 



per cent 

0.890 



prr cent 

0.117 



prr eenl 

13.14 



When the bran is digested in dilute acetic acid the hydrolysis 
is greater than with the same strength hydrochloric acid, as 
shown by the following results: 



SOLVENT 



0.2% acetic 
0.2% acetic 



TIME 

hr». 

3 

24 



TABLE VIL 

TOTAL 

PHOSPHORUS IN 

EXTRACT 

percent 

1.21 
1.23 



IHrOROANIC 

PHOSPHORUS IN 

EXTRACT 

per cent 

0.605 
1.050 



TOTAL 

PHOSPHORUS AS 

INORGANIC 



per cent 

57.03 
85.54 



The maximum activity of the enzyme, as shown by the great- 
est hydrolysis, was obtained by digesting bran for forty-eight 
hours in water and then adding 0.1 per cent hydrochloric acid 
and allowing it to stand for another twenty-four hours. Ten 
grams of bran were digested in 100 cc. of water for forty-eight 
hours and then 100 cc. of 0.2 per cent hydrochloric acid were 
added. After standing for twenty-four hours more it was filtered 
and 20 cc. of the filtrate were used for each determination. 





TABLE VIII. 




TOTAL PHOSPHORUS 


1 

INORGANIC PHOSPHORUS 


TOTAL PHOSPHORUS 


IN KXTRACT 


IN EXTRACT 


AH INORGANIC 


fM^ cent 


per cent 


per cent 


1.24 


1.14 


02.37 



490 Organic Phosphorus Compound of Wheat Bran 

Destruction of the enzyme phytase by heat. 

As is shown by the following experiments, the hydrolytic action 
of the enzjnne is completely destroyed when bran is exposed to 
the action of lx)iling water or boiling dilute hydrochloric acid for 
a short time. Ten grams of bran were placed in an Erlenmeyer 
flask, and 100 cc. of boiling water poured over it. The solution 
was then heated for a few minutes until the water toiled. It 
was allowed to digest for twenty-four hours at room tem[>erature. 
Another lot of bran \N'as treated in the same way, but with 0.2 
per cent hydrochloric acid instead of water. The resulting ex- 
tracts could not be filtered. The whole was therefore diluted 
with 100 cc. of 2 per cent hydrochloric acid and allowed to stand 
for a few minutes in order to settle. Lots of 20 cc. of the liquid 
were then taken for each detennination. 







TABLE IX. 










TOTAL 


INORGANIC 


TOTAL 


HOLVEXT 


TIME 


PHOflPHORUS IV 


PHOAPHORU8 IN 


PHOHPHORL'A A« 


f 


Afj*. 


EXTRACT 
per cent 


EXTRACT 
per cent 


INORGANIC 




jter cfnt 


Water 


24 


0.886 


0.108 


12. 6C 


0.2^;, HCl 


24 


1.140 


0.105 


9.22 



The hydrol>i;ic action of the enzyme is likewise destroyed 
by exposing bran for a short time to 0.5 per cent hydrochloric 
acid or by digesting bran in dilute ammonia. Ten grams of bran 
were digested in 100 cc. of 0.5 per cent hydrochloric acid for 
one and one-half hours. The acid was then nearly neutralized 
with ammonia, leaving the solution faintly acid, and then stand- 
ing for twenty-four hours. Ten grams of bran were digested in 
100 cc. of 0.25 per cent ammonia for twenty-four hours. It was 
then acidified witli dilute hydrochloric acid. 







TABLE X. 






1 




TOTAL 


INORGANIC 


TOTAL 


SOLVENT 


TIME 


, PHO.*«PHORL'K IN 


PHOaPHOHUS IN 


PHOSPIIOKUA A8 


■ 


hrs. 


1 EXTRACr ' 

1 per cent 


EXTRACT 

per ctnt 


INOROA.NIC 




jter cent 


0..5%IIC1 


24 


0.661 


0.115 


17.46 


0.25% NH, 


24 


0.922 


0.087 


9.51 



R. J, Anderson 491 

The results reported in this and the preceding paper will natu- 
rally change the interpretation of our previous investigations 
concerning the nature of the organic phosphorus compound of 
wheat bran. We have heretofore been under the impression that 
the substances which we have isolated from wheat bran, after 
digesting in 0.2 per cent hydrochloric acid, represented the com- 
pounds originally present in the bran. That this supposition is 
erroneous is evident, since the action of a bran extract on phytin 
gives rise to identical compounds. 

A consideration of the results reported in Table V likewise 
shows that there is a wide difference between the action of 0.2 
and 1.0 per cent hydrochloric acid on bran. In the first case the 
extract contains 0.925 per cent inorganic phosphorus, and in the 
second only 0.081 per cent. 

These facts lead to the conclusion that the products which 
we have previously isolated from wheat bran represent only such 
intennediate compounds as have been formed by the action of 
the enzyme phytase upon the original phytin of the bran. 

SUMMARY. 

The results herehi reported amplify and confirm the experi- 
ments of Suzuki, Yoshimura, and Takaishi, and of Plimmer con- 
cerning the presence of the enzyme phytase in wheat bran, which 
is capable of hydrolyzing the ph3rtin with the production of 
inorganic phosphoric acid. 

The maximum activity of the enzyme has been shown to occur 
in